Showing 1–47 of 47 results for author: Fuentes, M

Spacing Out: On the Reliability of Binaural Music Source Separation Metrics

Authors: Richa Namballa, Magdalena Fuentes

Abstract: Despite the rising popularity of immersive audio, binaural music remains underexplored in music information retrieval (MIR), particularly regarding the task of music source separation (MSS). While existing stereo MSS models can process binaural audio, they often degrade the spatial quality of the separated stems and undermine listener immersion. Through a perceptual study comparing binaural and st… ▽ More Despite the rising popularity of immersive audio, binaural music remains underexplored in music information retrieval (MIR), particularly regarding the task of music source separation (MSS). While existing stereo MSS models can process binaural audio, they often degrade the spatial quality of the separated stems and undermine listener immersion. Through a perceptual study comparing binaural and stereo MSS outputs, we evaluate how well objective spatial distortion metrics correlate with human perception. Our findings reveal varied agreement between these metrics and human judgment, highlighting a lack of reliability when used to evaluate binaural music tasks. Specifically, we find that Interaural Time Difference (ITD) estimation is highly sensitive to noise and separation artifacts. In evaluating two alternative ITD estimation methods, we uncover a critical trade-off between robustness and accuracy, particularly for narrow-band instruments like bass. These results underscore the need for accurate, interpretable spatial metrics designed for binaural music to develop models that preserve source localization and listener immersion. △ Less

Submitted 28 July, 2026; originally announced July 2026.

Comments: 6 pages + references, 6 figures, 1 table, 27th International Society for Music Information Retrieval (ISMIR) Conference

An Evaluation Framework for Structured Audio Captions Validated by Controlled Perturbations

Authors: Liang-Yuan Wu, Sripathi Sridhar, Mark Cartwright, Magdalena Fuentes

Abstract: Recent advancements in automated audio captioning (AAC) have shifted from monolithic sentence generation toward structured formats that explicitly disentangle distinct acoustic and semantic properties. However, evaluating this heterogeneous data remains a significant challenge. Existing caption metrics focus on flat textual outputs and fail to reliably assess multimodal attributes. To bridge this… ▽ More Recent advancements in automated audio captioning (AAC) have shifted from monolithic sentence generation toward structured formats that explicitly disentangle distinct acoustic and semantic properties. However, evaluating this heterogeneous data remains a significant challenge. Existing caption metrics focus on flat textual outputs and fail to reliably assess multimodal attributes. To bridge this gap, we propose a multi-axis evaluation framework tailored for structured audio descriptions. Building on the AudioCards dataset, we evaluate outputs across five orthogonal axes: tag-sets, descriptions, logical reasoning, numeric measurements, and spectral profiles. Our approach combines Large Language Model (LLM) judges to capture semantic nuance with deterministic computational metrics to precisely measure acoustic deviations. To rigorously validate the reliability of this framework, we introduce a controlled perturbation testing protocol that injects typed, graded errors into groundtruth annotations. Our results demonstrate that this framework successfully distinguishes meaning-preserving paraphrases from genuine semantic and acoustic corruptions. △ Less

Submitted 23 July, 2026; originally announced July 2026.

Comments: submitted to DCASE 2026

Deployment-Centered Evaluation: Predicting Query-Level Rejection Risk in a Clinical LLM System

Authors: Alyssa Unell, Miguel Fuentes, Brenna Li, Bridget Lin, Meena Jagadeesan, Sanmi Koyejo, Nigam Shah

Abstract: Large language models (LLMs) are increasingly integrated into clinical systems, making it essential to evaluate the real-world utility of these systems. However, static benchmarks tend to measure correctness rather than user acceptance, aggregate performance across queries, and require densely annotated datasets -- leading to major blind spots for evaluating clinical systems. In this work, we perf… ▽ More Large language models (LLMs) are increasingly integrated into clinical systems, making it essential to evaluate the real-world utility of these systems. However, static benchmarks tend to measure correctness rather than user acceptance, aggregate performance across queries, and require densely annotated datasets -- leading to major blind spots for evaluating clinical systems. In this work, we perform a deployment-centered evaluation of an LLM system embedded within electronic health records at an academic medical center, where user feedback is sparse but closely reflects the deployment conditions. Specifically, we train a pre-response classifier that estimates the risk that a future interaction will result in the user rejecting the LLM response, based on query content and deployment-specific context available before generation. We conduct a prospective analysis of our model over 4.5 months of user feedback, finding that our prediction model achieves an AUROC of 0.719. Further, we estimate the benefit of such predictions in two downstream use cases (guardrail triggering and abstention). Our key conceptual insight is that making use of deployment-specific context (i.e., the provider type, department name, language model used for response), as opposed to only query content, improves the ability to predict whether the user will reject the system output. Altogether, our empirical case study demonstrates the feasibility of predicting user rejection using deployment-specific context, opening the door to targeted guardrails. △ Less

Submitted 10 June, 2026; originally announced June 2026.

Private Adaptive Covariance Estimation via Gaussian Graphical Models

Authors: Cecilia Ferrando, Miguel Fuentes, Brett Mullins, Cameron Musco, Daniel Sheldon

Abstract: We propose PACE-GGM, a data-adaptive differentially private method for covariance estimation that concentrates its privacy budget on the most informative entries of the empirical covariance matrix, rather than perturbing all entries. This applies in the natural setting where the modeler supplies separate bounds for each variable, so that individual entries can be measured with less noise than the… ▽ More We propose PACE-GGM, a data-adaptive differentially private method for covariance estimation that concentrates its privacy budget on the most informative entries of the empirical covariance matrix, rather than perturbing all entries. This applies in the natural setting where the modeler supplies separate bounds for each variable, so that individual entries can be measured with less noise than the full matrix. In each round, our method selects a poorly approximated entry, measures it using the Gaussian mechanism, and then reconstructs a full covariance matrix using a maximum-entropy reconstruction objective, leading to a Gaussian graphical model structure. Experiments on diverse real-world datasets demonstrate consistent improvements in estimation error with respect to the Gaussian mechanism and other baselines, particularly in high-dimensional and low-to-moderate privacy regimes. △ Less

Submitted 22 May, 2026; originally announced May 2026.

Fast Private Adaptive Query Answering for Large Data Domains

Authors: Miguel Fuentes, Brett Mullins, Yingtai Xiao, Daniel Kifer, Cameron Musco, Daniel Sheldon

Abstract: Privately releasing marginals of a tabular dataset is a foundational problem in differential privacy. However, state-of-the-art mechanisms suffer from a computational bottleneck when marginal estimates are reconstructed from noisy measurements. Recently, residual queries were introduced and shown to lead to highly efficient reconstruction in the batch query answering setting. We introduce new tech… ▽ More Privately releasing marginals of a tabular dataset is a foundational problem in differential privacy. However, state-of-the-art mechanisms suffer from a computational bottleneck when marginal estimates are reconstructed from noisy measurements. Recently, residual queries were introduced and shown to lead to highly efficient reconstruction in the batch query answering setting. We introduce new techniques to integrate residual queries into state-of-the-art adaptive mechanisms such as AIM. Our contributions include a novel conceptual framework for residual queries using multi-dimensional arrays, lazy updating strategies, and adaptive optimization of the per-round privacy budget allocation. Together these contributions reduce error, improve speed, and simplify residual query operations. We integrate these innovations into a new mechanism (AIM+GReM), which improves AIM by using fast residual-based reconstruction instead of a graphical model approach. Our mechanism is orders of magnitude faster than the original framework and demonstrates competitive error and greatly improved scalability. △ Less

Submitted 25 February, 2026; v1 submitted 5 February, 2026; originally announced February 2026.

Adoption and Use of LLMs at an Academic Medical Center

Authors: Nigam H. Shah, Nerissa Ambers, Abby Pandya, Timothy Keyes, Juan M. Banda, Srikar Nallan, Carlene Lugtu, Artem A. Trotsyuk, Suhana Bedi, Alyssa Unell, Miguel Fuentes, Francois Grolleau, Sneha S. Jain, Jonathan Chen, Devdutta Dash, Danton Char, Aditya Sharma, Duncan McElfresh, Patrick Scully, Vishanthan Kumar, Clancy Dennis, Connor OBrien, Satchi Mouniswamy, Elvis Jones, Krishna Jasti , et al. (33 additional authors not shown)

Abstract: While large language models (LLMs) can support clinical documentation needs, standalone tools struggle with "workflow friction" from manual data entry. We developed ChatEHR, a system that enables the use of LLMs with the entire patient timeline spanning several years. ChatEHR enables automations - which are static combinations of prompts and data that perform a fixed task - and interactive use in… ▽ More While large language models (LLMs) can support clinical documentation needs, standalone tools struggle with "workflow friction" from manual data entry. We developed ChatEHR, a system that enables the use of LLMs with the entire patient timeline spanning several years. ChatEHR enables automations - which are static combinations of prompts and data that perform a fixed task - and interactive use in the electronic health record (EHR) via a user interface (UI). The resulting ability to sift through patient medical records for diverse use-cases such as pre-visit chart review, screening for transfer eligibility, monitoring for surgical site infections, and chart abstraction, redefines LLM use as an institutional capability. This system, accessible after user-training, enables continuous monitoring and evaluation of LLM use. In 1.5 years, we built 7 automations and 1075 users have trained to become routine users of the UI, engaging in 23,000 sessions in the first 3 months of launch. For automations, being model-agnostic and accessing multiple types of data was essential for matching specific clinical or administrative tasks with the most appropriate LLM. Benchmark-based evaluations proved insufficient for monitoring and evaluation of the UI, requiring new methods to monitor performance. Generation of summaries was the most frequent task in the UI, with an estimated 0.73 hallucinations and 1.60 inaccuracies per generation. The resulting mix of cost savings, time savings, and revenue growth required a value assessment framework to prioritize work as well as quantify the impact of using LLMs. Initial estimates are $6M savings in the first year of use, without quantifying the benefit of the better care offered. Such a "build-from-within" strategy provides an opportunity for health systems to maintain agency via a vendor-agnostic, internally governed LLM platform. △ Less

Submitted 30 April, 2026; v1 submitted 20 January, 2026; originally announced February 2026.

Training-Free Adaptation of New-Generation LLMs using Legacy Clinical Models

Authors: Sasha Ronaghi, Chloe Stanwyck, Asad Aali, Amir Ronaghi, Miguel Fuentes, Tina Hernandez-Boussard, Emily Alsentzer

Abstract: Adapting language models to the clinical domain through continued pretraining and instruction tuning requires costly retraining for each new model generation. We propose Cross-Architecture Proxy Tuning (CAPT), a model-ensembling approach that enables training-free adaptation of state-of-the-art general-domain models using existing clinical models. CAPT supports models with disjoint vocabularies, l… ▽ More Adapting language models to the clinical domain through continued pretraining and instruction tuning requires costly retraining for each new model generation. We propose Cross-Architecture Proxy Tuning (CAPT), a model-ensembling approach that enables training-free adaptation of state-of-the-art general-domain models using existing clinical models. CAPT supports models with disjoint vocabularies, leveraging contrastive decoding to selectively inject clinically relevant signals while preserving the general-domain model's reasoning and fluency. On six clinical classification and text-generation tasks, CAPT with a new-generation general-domain model and an older-generation clinical model consistently outperforms both models individually and state-of-the-art ensembling approaches (average +17.6\% over UniTE, +41.4\% over proxy tuning across tasks). Through token-level analysis and physician case studies, we demonstrate that CAPT amplifies clinically actionable language, reduces context errors, and increases clinical specificity. This technique especially benefits healthcare institutions with constrained computational capacity that cannot support iterative clinical training and want to adopt emerging general-domain model advances. △ Less

Submitted 28 April, 2026; v1 submitted 6 January, 2026; originally announced January 2026.

Journal ref: Proceedings of the 7th Conference on Health, Inference, and Learning, PMLR 333:354-388, 2026

Monitoring Deployed AI Systems in Health Care

Authors: Timothy Keyes, Alison Callahan, Abby S. Pandya, Nerissa Ambers, Juan M. Banda, Miguel Fuentes, Carlene Lugtu, Pranav Masariya, Srikar Nallan, Connor O'Brien, Thomas Wang, Emily Alsentzer, Jonathan H. Chen, Dev Dash, Matthew A. Eisenberg, Patricia Garcia, Nikesh Kotecha, Anurang Revri, Michael A. Pfeffer, Nigam H. Shah, Sneha S. Jain

Abstract: Post-deployment monitoring of artificial intelligence (AI) systems in health care is essential to ensure their safety, quality, and sustained benefit-and to support governance decisions about which systems to update, modify, or decommission. Motivated by these needs, we developed a framework for monitoring deployed AI systems grounded in the mandate to take specific actions when they fail to behav… ▽ More Post-deployment monitoring of artificial intelligence (AI) systems in health care is essential to ensure their safety, quality, and sustained benefit-and to support governance decisions about which systems to update, modify, or decommission. Motivated by these needs, we developed a framework for monitoring deployed AI systems grounded in the mandate to take specific actions when they fail to behave as intended. This framework, which is now actively used at Stanford Health Care, is organized around three complementary principles: system integrity, performance, and impact. System integrity monitoring focuses on maximizing system uptime, detecting runtime errors, and identifying when changes to the surrounding IT ecosystem have unintended effects. Performance monitoring focuses on maintaining accurate system behavior in the face of changing health care practices (and thus input data) over time. Impact monitoring assesses whether a deployed system continues to have value in the form of benefit to clinicians and patients. Drawing on examples of deployed AI systems at our academic medical center, we provide practical guidance for creating monitoring plans based on these principles that specify which metrics to measure, when those metrics should be reviewed, who is responsible for acting when metrics change, and what concrete follow-up actions should be taken-for both traditional and generative AI. We also discuss challenges to implementing this framework, including the effort and cost of monitoring for health systems with limited resources and the difficulty of incorporating data-driven monitoring practices into complex organizations where conflicting priorities and definitions of success often coexist. This framework offers a practical template and starting point for health systems seeking to ensure that AI deployments remain safe and effective over time. △ Less

Submitted 15 January, 2026; v1 submitted 9 December, 2025; originally announced December 2025.

Comments: 36 pages, 3 figures

Structured Prompts Improve Evaluation of Language Models

Authors: Asad Aali, Muhammad Ahmed Mohsin, Vasiliki Bikia, Arnav Singhvi, Richard Gaus, Suhana Bedi, Hejie Cui, Miguel Fuentes, Alyssa Unell, Yifan Mai, Jordan Cahoon, Michael Pfeffer, Roxana Daneshjou, Sanmi Koyejo, Emily Alsentzer, Christopher Potts, Nigam H. Shah, Akshay S. Chaudhari

Abstract: As language models (LMs) are increasingly adopted across domains, high-quality benchmarking frameworks are essential for guiding deployment decisions. In practice, however, frameworks such as Holistic Evaluation of Language Models (HELM) typically evaluate models under a single static prompt configuration, even though model behavior depends strongly on prompt choice. As a result, reported scores c… ▽ More As language models (LMs) are increasingly adopted across domains, high-quality benchmarking frameworks are essential for guiding deployment decisions. In practice, however, frameworks such as Holistic Evaluation of Language Models (HELM) typically evaluate models under a single static prompt configuration, even though model behavior depends strongly on prompt choice. As a result, reported scores can reflect prompt choice as much as model capability. Declarative prompting frameworks such as DSPy offer a scalable way to evaluate models under a set of structured prompting strategies rather than a static prompt configuration. We present a reproducible DSPy+HELM framework for studying how prompt choice impacts reported benchmark outcomes. Using five prompting methods, we evaluate four frontier and two open-source LMs across seven benchmarks against existing HELM baseline scores. By evaluating LMs across a family of prompt configurations, we find that prompt choice can materially impact leaderboard outcomes. In particular, structured prompting improves performance (by 6% on average), alters comparisons (leaderboard rankings shift on 5/7 benchmarks), with most gains coming from introducing chain-of-thought, and little additional benefit from more advanced optimizers. To our knowledge, this is the first study to systematically integrate structured prompting into an established evaluation framework and quantify how prompt choice alone can impact benchmark conclusions. We open-source (i) DSPy+HELM Evaluation (https://github.com/stanford-crfm/helm/pull/3893) and (ii) Prompt Optimization Pipeline (https://github.com/StanfordMIMI/dspy-helm). △ Less

Submitted 1 April, 2026; v1 submitted 25 November, 2025; originally announced November 2025.

Twenty-Five Years of MIR Research: Achievements, Practices, Evaluations, and Future Challenges

Authors: Geoffroy Peeters, Zafar Rafii, Magdalena Fuentes, Zhiyao Duan, Emmanouil Benetos, Juhan Nam, Yuki Mitsufuji

Abstract: In this paper, we trace the evolution of Music Information Retrieval (MIR) over the past 25 years. While MIR gathers all kinds of research related to music informatics, a large part of it focuses on signal processing techniques for music data, fostering a close relationship with the IEEE Audio and Acoustic Signal Processing Technical Commitee. In this paper, we reflect the main research achievemen… ▽ More In this paper, we trace the evolution of Music Information Retrieval (MIR) over the past 25 years. While MIR gathers all kinds of research related to music informatics, a large part of it focuses on signal processing techniques for music data, fostering a close relationship with the IEEE Audio and Acoustic Signal Processing Technical Commitee. In this paper, we reflect the main research achievements of MIR along the three EDICS related to music analysis, processing and generation. We then review a set of successful practices that fuel the rapid development of MIR research. One practice is the annual research benchmark, the Music Information Retrieval Evaluation eXchange, where participants compete on a set of research tasks. Another practice is the pursuit of reproducible and open research. The active engagement with industry research and products is another key factor for achieving large societal impacts and motivating younger generations of students to join the field. Last but not the least, the commitment to diversity, equity and inclusion ensures MIR to be a vibrant and open community where various ideas, methodologies, and career pathways collide. We finish by providing future challenges MIR will have to face. △ Less

Submitted 10 November, 2025; originally announced November 2025.

Latent Multi-view Learning for Robust Environmental Sound Representations

Authors: Sivan Ding, Julia Wilkins, Magdalena Fuentes, Juan Pablo Bello

Abstract: Self-supervised learning (SSL) approaches, such as contrastive and generative methods, have advanced environmental sound representation learning using unlabeled data. However, how these approaches can complement each other within a unified framework remains relatively underexplored. In this work, we propose a multi-view learning framework that integrates contrastive principles into a generative pi… ▽ More Self-supervised learning (SSL) approaches, such as contrastive and generative methods, have advanced environmental sound representation learning using unlabeled data. However, how these approaches can complement each other within a unified framework remains relatively underexplored. In this work, we propose a multi-view learning framework that integrates contrastive principles into a generative pipeline to capture sound source and device information. Our method encodes compressed audio latents into view-specific and view-common subspaces, guided by two self-supervised objectives: contrastive learning for targeted information flow between subspaces, and reconstruction for overall information preservation. We evaluate our method on an urban sound sensor network dataset for sound source and sensor classification, demonstrating improved downstream performance over traditional SSL techniques. Additionally, we investigate the model's potential to disentangle environmental sound attributes within the structured latent space under varied training configurations. △ Less

Submitted 27 October, 2025; v1 submitted 2 October, 2025; originally announced October 2025.

Comments: Accepted to DCASE 2025 Workshop. 4+1 pages, 2 figures, 2 tables

Post-Training Quantization for Audio Diffusion Transformers

Authors: Tanmay Khandelwal, Magdalena Fuentes

Abstract: Diffusion Transformers (DiTs) enable high-quality audio synthesis but are often computationally intensive and require substantial storage, which limits their practical deployment. In this paper, we present a comprehensive evaluation of post-training quantization (PTQ) techniques for audio DiTs, analyzing the trade-offs between static and dynamic quantization schemes. We explore two practical exten… ▽ More Diffusion Transformers (DiTs) enable high-quality audio synthesis but are often computationally intensive and require substantial storage, which limits their practical deployment. In this paper, we present a comprehensive evaluation of post-training quantization (PTQ) techniques for audio DiTs, analyzing the trade-offs between static and dynamic quantization schemes. We explore two practical extensions (1) a denoising-timestep-aware smoothing method that adapts quantization scales per-input-channel and timestep to mitigate activation outliers, and (2) a lightweight low-rank adapter (LoRA)-based branch derived from singular value decomposition (SVD) to compensate for residual weight errors. Using Stable Audio Open we benchmark W8A8 and W4A8 configurations across objective metrics and human perceptual ratings. Our results show that dynamic quantization preserves fidelity even at lower precision, while static methods remain competitive with lower latency. Overall, our findings show that low-precision DiTs can retain high-fidelity generation while reducing memory usage by up to 79%. △ Less

Submitted 30 September, 2025; originally announced October 2025.

Comments: 5 pages, 4 figures, accepted at IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) 2025

Investigating Modality Contribution in Audio LLMs for Music

Authors: Giovana Morais, Magdalena Fuentes

Abstract: Audio Large Language Models (Audio LLMs) enable human-like conversation about music, yet it is unclear if they are truly listening to the audio or just using textual reasoning, as recent benchmarks suggest. This paper investigates this issue by quantifying the contribution of each modality to a model's output. We adapt the MM-SHAP framework, a performance-agnostic score based on Shapley values tha… ▽ More Audio Large Language Models (Audio LLMs) enable human-like conversation about music, yet it is unclear if they are truly listening to the audio or just using textual reasoning, as recent benchmarks suggest. This paper investigates this issue by quantifying the contribution of each modality to a model's output. We adapt the MM-SHAP framework, a performance-agnostic score based on Shapley values that quantifies the relative contribution of each modality to a model's prediction. We evaluate two models on the MuChoMusic benchmark and find that the model with higher accuracy relies more on text to answer questions, but further inspection shows that even if the overall audio contribution is low, models can successfully localize key sound events, suggesting that audio is not entirely ignored. Our study is the first application of MM-SHAP to Audio LLMs and we hope it will serve as a foundational step for future research in explainable AI and audio. △ Less

Submitted 2 February, 2026; v1 submitted 24 September, 2025; originally announced September 2025.

Comments: 5 pages, 2 figures, accepted at ICASSP 2026

Learning from Silence and Noise for Visual Sound Source Localization

Authors: Xavier Juanola, Giovana Morais, Magdalena Fuentes, Gloria Haro

Abstract: Visual sound source localization is a fundamental perception task that aims to detect the location of sounding sources in a video given its audio. Despite recent progress, we identify two shortcomings in current methods: 1) most approaches perform poorly in cases with low audio-visual semantic correspondence such as silence, noise, and offscreen sounds, i.e. in the presence of negative audio; and… ▽ More Visual sound source localization is a fundamental perception task that aims to detect the location of sounding sources in a video given its audio. Despite recent progress, we identify two shortcomings in current methods: 1) most approaches perform poorly in cases with low audio-visual semantic correspondence such as silence, noise, and offscreen sounds, i.e. in the presence of negative audio; and 2) most prior evaluations are limited to positive cases, where both datasets and metrics convey scenarios with a single visible sound source in the scene. To address this, we introduce three key contributions. First, we propose a new training strategy that incorporates silence and noise, which improves performance in positive cases, while being more robust against negative sounds. Our resulting self-supervised model, SSL-SaN, achieves state-of-the-art performance compared to other self-supervised models, both in sound localization and cross-modal retrieval. Second, we propose a new metric that quantifies the trade-off between alignment and separability of auditory and visual features across positive and negative audio-visual pairs. Third, we present IS3+, an extended and improved version of the IS3 synthetic dataset with negative audio. Our data, metrics and code are available on the https://xavijuanola.github.io/SSL-SaN/. △ Less

Submitted 29 August, 2025; originally announced August 2025.

Comments: 10 pages, 2 figures, 4 tables + Supplementary Material

Live Vocal Extraction from K-pop Performances

Authors: Yujin Kim, Richa Namballa, Magdalena Fuentes

Abstract: K-pop's global success is fueled by its dynamic performances and vibrant fan engagement. Inspired by K-pop fan culture, we propose a methodology for automatically extracting live vocals from performances. We use a combination of source separation, cross-correlation, and amplitude scaling to automatically remove pre-recorded vocals and instrumentals from a live performance. Our preliminary work int… ▽ More K-pop's global success is fueled by its dynamic performances and vibrant fan engagement. Inspired by K-pop fan culture, we propose a methodology for automatically extracting live vocals from performances. We use a combination of source separation, cross-correlation, and amplitude scaling to automatically remove pre-recorded vocals and instrumentals from a live performance. Our preliminary work introduces the task of live vocal separation and provides a foundation for future research in this topic. △ Less

Submitted 27 August, 2025; originally announced August 2025.

Comments: 2 pages + references, 1 figure, Extended Abstracts for the Late-Breaking Demo Session of the 26th International Society for Music Information Retrieval Conference

VLH: Vision-Language-Haptics Foundation Model

Authors: Luis Francisco Moreno Fuentes, Muhammad Haris Khan, Miguel Altamirano Cabrera, Valerii Serpiva, Dmitri Iarchuk, Yara Mahmoud, Issatay Tokmurziyev, Dzmitry Tsetserukou

Abstract: We present VLH, a novel Visual-Language-Haptic Foundation Model that unifies perception, language, and tactile feedback in aerial robotics and virtual reality. Unlike prior work that treats haptics as a secondary, reactive channel, VLH synthesizes mid-air force and vibration cues as a direct consequence of contextual visual understanding and natural language commands. Our platform comprises an 8-i… ▽ More We present VLH, a novel Visual-Language-Haptic Foundation Model that unifies perception, language, and tactile feedback in aerial robotics and virtual reality. Unlike prior work that treats haptics as a secondary, reactive channel, VLH synthesizes mid-air force and vibration cues as a direct consequence of contextual visual understanding and natural language commands. Our platform comprises an 8-inch quadcopter equipped with dual inverse five-bar linkage arrays for localized haptic actuation, an egocentric VR camera, and an exocentric top-down view. Visual inputs and language instructions are processed by a fine-tuned OpenVLA backbone - adapted via LoRA on a bespoke dataset of 450 multimodal scenarios - to output a 7-dimensional action vector (Vx, Vy, Vz, Hx, Hy, Hz, Hv). INT8 quantization and a high-performance server ensure real-time operation at 4-5 Hz. In human-robot interaction experiments (90 flights), VLH achieved a 56.7% success rate for target acquisition (mean reach time 21.3 s, pose error 0.24 m) and 100% accuracy in texture discrimination. Generalization tests yielded 70.0% (visual), 54.4% (motion), 40.0% (physical), and 35.0% (semantic) performance on novel tasks. These results demonstrate VLH's ability to co-evolve haptic feedback with perceptual reasoning and intent, advancing expressive, immersive human-robot interactions. △ Less

Submitted 2 August, 2025; originally announced August 2025.

Balancing Information Preservation and Disentanglement in Self-Supervised Music Representation Learning

Authors: Julia Wilkins, Sivan Ding, Magdalena Fuentes, Juan Pablo Bello

Abstract: Recent advances in self-supervised learning (SSL) methods offer a range of strategies for capturing useful representations from music audio without the need for labeled data. While some techniques focus on preserving comprehensive details through reconstruction, others favor semantic structure via contrastive objectives. Few works examine the interaction between these paradigms in a unified SSL fr… ▽ More Recent advances in self-supervised learning (SSL) methods offer a range of strategies for capturing useful representations from music audio without the need for labeled data. While some techniques focus on preserving comprehensive details through reconstruction, others favor semantic structure via contrastive objectives. Few works examine the interaction between these paradigms in a unified SSL framework. In this work, we propose a multi-view SSL framework for disentangling music audio representations that combines contrastive and reconstructive objectives. The architecture is designed to promote both information fidelity and structured semantics of factors in disentangled subspaces. We perform an extensive evaluation on the design choices of contrastive strategies using music audio representations in a controlled setting. We find that while reconstruction and contrastive strategies exhibit consistent trade-offs, when combined effectively, they complement each other; this enables the disentanglement of music attributes without compromising information integrity. △ Less

Submitted 30 July, 2025; originally announced July 2025.

Comments: In proceedings of WASPAA 2025. 4 pages, 4 figures, 1 table

Do Music Source Separation Models Preserve Spatial Information in Binaural Audio?

Authors: Richa Namballa, Agnieszka Roginska, Magdalena Fuentes

Abstract: Binaural audio remains underexplored within the music information retrieval community. Motivated by the rising popularity of virtual and augmented reality experiences as well as potential applications to accessibility, we investigate how well existing music source separation (MSS) models perform on binaural audio. Although these models process two-channel inputs, it is unclear how effectively they… ▽ More Binaural audio remains underexplored within the music information retrieval community. Motivated by the rising popularity of virtual and augmented reality experiences as well as potential applications to accessibility, we investigate how well existing music source separation (MSS) models perform on binaural audio. Although these models process two-channel inputs, it is unclear how effectively they retain spatial information. In this work, we evaluate how several popular MSS models preserve spatial information on both standard stereo and novel binaural datasets. Our binaural data is synthesized using stems from MUSDB18-HQ and open-source head-related transfer functions by positioning instrument sources randomly along the horizontal plane. We then assess the spatial quality of the separated stems using signal processing and interaural cue-based metrics. Our results show that stereo MSS models fail to preserve the spatial information critical for maintaining the immersive quality of binaural audio, and that the degradation depends on model architecture as well as the target instrument. Finally, we highlight valuable opportunities for future work at the intersection of MSS and immersive audio. △ Less

Submitted 30 June, 2025; originally announced July 2025.

Comments: 6 pages + references, 4 figures, 2 tables, 26th International Society for Music Information Retrieval (ISMIR) Conference

Domain Adaptation Method and Modality Gap Impact in Audio-Text Models for Prototypical Sound Classification

Authors: Emiliano Acevedo, Martín Rocamora, Magdalena Fuentes

Abstract: Audio-text models are widely used in zero-shot environmental sound classification as they alleviate the need for annotated data. However, we show that their performance severely drops in the presence of background sound sources. Our analysis reveals that this degradation is primarily driven by SNR levels of background soundscapes, and independent of background type. To address this, we propose a n… ▽ More Audio-text models are widely used in zero-shot environmental sound classification as they alleviate the need for annotated data. However, we show that their performance severely drops in the presence of background sound sources. Our analysis reveals that this degradation is primarily driven by SNR levels of background soundscapes, and independent of background type. To address this, we propose a novel method that quantifies and integrates the contribution of background sources into the classification process, improving performance without requiring model retraining. Our domain adaptation technique enhances accuracy across various backgrounds and SNR conditions. Moreover, we analyze the modality gap between audio and text embeddings, showing that narrowing this gap improves classification performance. The method generalizes effectively across state-of-the-art prototypical approaches, showcasing its scalability and robustness for diverse environments. △ Less

Submitted 4 June, 2025; originally announced June 2025.

Comments: Accepted at INTERSPEECH 2025

LPASS: Linear Probes as Stepping Stones for vulnerability detection using compressed LLMs

Authors: Luis Ibanez-Lissen, Lorena Gonzalez-Manzano, Jose Maria de Fuentes, Nicolas Anciaux

Abstract: Large Language Models (LLMs) are being extensively used for cybersecurity purposes. One of them is the detection of vulnerable codes. For the sake of efficiency and effectiveness, compression and fine-tuning techniques are being developed, respectively. However, they involve spending substantial computational efforts. In this vein, we analyse how Linear Probes (LPs) can be used to provide an estim… ▽ More Large Language Models (LLMs) are being extensively used for cybersecurity purposes. One of them is the detection of vulnerable codes. For the sake of efficiency and effectiveness, compression and fine-tuning techniques are being developed, respectively. However, they involve spending substantial computational efforts. In this vein, we analyse how Linear Probes (LPs) can be used to provide an estimation on the performance of a compressed LLM at an early phase -- before fine-tuning. We also show their suitability to set the cut-off point when applying layer pruning compression. Our approach, dubbed $LPASS$, is applied in BERT and Gemma for the detection of 12 of MITRE's Top 25 most dangerous vulnerabilities on 480k C/C++ samples. LPs can be computed in 142.97 s. and provide key findings: (1) 33.3 \% and 72.2\% of layers can be removed, respectively, with no precision loss; (2) they provide an early estimate of the post-fine-tuning and post-compression model effectiveness, with 3\% and 8.68\% as the lowest and average precision errors, respectively. $LPASS$-based LLMs outperform the state of the art, reaching 86.9\% of accuracy in multi-class vulnerability detection. Interestingly, $LPASS$-based compressed versions of Gemma outperform the original ones by 1.6\% of F1-score at a maximum while saving 29.4 \% and 23.8\% of training and inference time and 42.98\% of model size. △ Less

Submitted 30 May, 2025; originally announced May 2025.

MedHELM: Holistic Evaluation of Large Language Models for Medical Tasks

Authors: Suhana Bedi, Hejie Cui, Miguel Fuentes, Alyssa Unell, Michael Wornow, Juan M. Banda, Nikesh Kotecha, Timothy Keyes, Yifan Mai, Mert Oez, Hao Qiu, Shrey Jain, Leonardo Schettini, Mehr Kashyap, Jason Alan Fries, Akshay Swaminathan, Philip Chung, Fateme Nateghi, Asad Aali, Ashwin Nayak, Shivam Vedak, Sneha S. Jain, Birju Patel, Oluseyi Fayanju, Shreya Shah , et al. (56 additional authors not shown)

Abstract: While large language models (LLMs) achieve near-perfect scores on medical licensing exams, these evaluations inadequately reflect the complexity and diversity of real-world clinical practice. We introduce MedHELM, an extensible evaluation framework for assessing LLM performance for medical tasks with three key contributions. First, a clinician-validated taxonomy spanning 5 categories, 22 subcatego… ▽ More While large language models (LLMs) achieve near-perfect scores on medical licensing exams, these evaluations inadequately reflect the complexity and diversity of real-world clinical practice. We introduce MedHELM, an extensible evaluation framework for assessing LLM performance for medical tasks with three key contributions. First, a clinician-validated taxonomy spanning 5 categories, 22 subcategories, and 121 tasks developed with 29 clinicians. Second, a comprehensive benchmark suite comprising 35 benchmarks (17 existing, 18 newly formulated) providing complete coverage of all categories and subcategories in the taxonomy. Third, a systematic comparison of LLMs with improved evaluation methods (using an LLM-jury) and a cost-performance analysis. Evaluation of 9 frontier LLMs, using the 35 benchmarks, revealed significant performance variation. Advanced reasoning models (DeepSeek R1: 66% win-rate; o3-mini: 64% win-rate) demonstrated superior performance, though Claude 3.5 Sonnet achieved comparable results at 40% lower estimated computational cost. On a normalized accuracy scale (0-1), most models performed strongly in Clinical Note Generation (0.73-0.85) and Patient Communication & Education (0.78-0.83), moderately in Medical Research Assistance (0.65-0.75), and generally lower in Clinical Decision Support (0.56-0.72) and Administration & Workflow (0.53-0.63). Our LLM-jury evaluation method achieved good agreement with clinician ratings (ICC = 0.47), surpassing both average clinician-clinician agreement (ICC = 0.43) and automated baselines including ROUGE-L (0.36) and BERTScore-F1 (0.44). Claude 3.5 Sonnet achieved comparable performance to top models at lower estimated cost. These findings highlight the importance of real-world, task-specific evaluation for medical use of LLMs and provides an open source framework to enable this. △ Less

Submitted 2 June, 2025; v1 submitted 26 May, 2025; originally announced May 2025.

Musical Source Separation of Brazilian Percussion

Authors: Richa Namballa, Giovana Morais, Magdalena Fuentes

Abstract: Musical source separation (MSS) has recently seen a big breakthrough in separating instruments from a mixture in the context of Western music, but research on non-Western instruments is still limited due to a lack of data. In this demo, we use an existing dataset of Brazilian sama percussion to create artificial mixtures for training a U-Net model to separate the surdo drum, a traditional instrume… ▽ More Musical source separation (MSS) has recently seen a big breakthrough in separating instruments from a mixture in the context of Western music, but research on non-Western instruments is still limited due to a lack of data. In this demo, we use an existing dataset of Brazilian sama percussion to create artificial mixtures for training a U-Net model to separate the surdo drum, a traditional instrument in samba. Despite limited training data, the model effectively isolates the surdo, given the drum's repetitive patterns and its characteristic low-pitched timbre. These results suggest that MSS systems can be successfully harnessed to work in more culturally-inclusive scenarios without the need of collecting extensive amounts of data. △ Less

Submitted 6 March, 2025; originally announced March 2025.

Comments: 2 pages + references, 1 figure, 1 table, Extended Abstracts for the Late-Breaking Demo Session of the 25th International Society for Music Information Retrieval Conference

Skip That Beat: Augmenting Meter Tracking Models for Underrepresented Time Signatures

Authors: Giovana Morais, Brian McFee, Magdalena Fuentes

Abstract: Beat and downbeat tracking models are predominantly developed using datasets with music in 4/4 meter, which decreases their generalization to repertories in other time signatures, such as Brazilian samba which is in 2/4. In this work, we propose a simple augmentation technique to increase the representation of time signatures beyond 4/4, namely 2/4 and 3/4. Our augmentation procedure works by remo… ▽ More Beat and downbeat tracking models are predominantly developed using datasets with music in 4/4 meter, which decreases their generalization to repertories in other time signatures, such as Brazilian samba which is in 2/4. In this work, we propose a simple augmentation technique to increase the representation of time signatures beyond 4/4, namely 2/4 and 3/4. Our augmentation procedure works by removing beat intervals from 4/4 annotated tracks. We show that the augmented data helps to improve downbeat tracking for underrepresented meters while preserving the overall performance of beat tracking in two different models. We also show that this technique helps improve downbeat tracking in an unseen samba dataset. △ Less

Submitted 18 February, 2025; originally announced February 2025.

Comments: 4 pages + references, 3 figures, 1st Latin American Music Information Retrieval (LAMIR) workshop

Strassen Attention, Split VC Dimension and Compositionality in Transformers

Authors: Alexander Kozachinskiy, Felipe Urrutia, Hector Jimenez, Tomasz Steifer, Germán Pizarro, Matías Fuentes, Francisco Meza, Cristian B. Calderon, Cristóbal Rojas

Abstract: We propose the first method to show theoretical limitations for one-layer softmax transformers with arbitrarily many precision bits (even infinite). We establish those limitations for three tasks that require advanced reasoning. The first task, Match 3 (Sanford et al., 2023), requires looking at all possible token triplets in an input sequence. The second and third tasks address compositionality-b… ▽ More We propose the first method to show theoretical limitations for one-layer softmax transformers with arbitrarily many precision bits (even infinite). We establish those limitations for three tasks that require advanced reasoning. The first task, Match 3 (Sanford et al., 2023), requires looking at all possible token triplets in an input sequence. The second and third tasks address compositionality-based reasoning: function composition (Peng et al., 2024) and binary relations composition, respectively. We formally prove the inability of one-layer softmax Transformers to solve any of these tasks. To overcome these limitations, we introduce Strassen attention and prove that, equipped with this mechanism, a one-layer transformer can in principle solve all these tasks. Importantly, we show that it enjoys sub-cubic running-time complexity, making it more scalable than similar previously proposed mechanisms, such as higher-order attention (Sanford et al., 2023). To complement our theoretical findings, we experimentally studied Strassen attention and compared it against standard (Vaswani et al, 2017), higher-order attention (Sanford et al., 2023), and triangular attention (Bergen et al. 2021). Our results help to disentangle all these attention mechanisms, highlighting their strengths and limitations. In particular, Strassen attention outperforms standard attention significantly on all the tasks. Altogether, understanding the theoretical limitations can guide research towards scalable attention mechanisms that improve the reasoning abilities of Transformers. △ Less

Submitted 25 September, 2025; v1 submitted 31 January, 2025; originally announced January 2025.

Comments: Accepted at NeurIPS 2025

LUMIA: Linear probing for Unimodal and MultiModal Membership Inference Attacks leveraging internal LLM states

Authors: Luis Ibanez-Lissen, Lorena Gonzalez-Manzano, Jose Maria de Fuentes, Nicolas Anciaux, Joaquin Garcia-Alfaro

Abstract: Large Language Models (LLMs) are increasingly used in a variety of applications, but concerns around membership inference have grown in parallel. Previous efforts focus on black-to-grey-box models, thus neglecting the potential benefit from internal LLM information. To address this, we propose the use of Linear Probes (LPs) as a method to detect Membership Inference Attacks (MIAs) by examining int… ▽ More Large Language Models (LLMs) are increasingly used in a variety of applications, but concerns around membership inference have grown in parallel. Previous efforts focus on black-to-grey-box models, thus neglecting the potential benefit from internal LLM information. To address this, we propose the use of Linear Probes (LPs) as a method to detect Membership Inference Attacks (MIAs) by examining internal activations of LLMs. Our approach, dubbed LUMIA, applies LPs layer-by-layer to get fine-grained data on the model inner workings. We test this method across several model architectures, sizes and datasets, including unimodal and multimodal tasks. In unimodal MIA, LUMIA achieves an average gain of 15.71 % in Area Under the Curve (AUC) over previous techniques. Remarkably, LUMIA reaches AUC>60% in 65.33% of cases -- an increment of 46.80% against the state of the art. Furthermore, our approach reveals key insights, such as the model layers where MIAs are most detectable. In multimodal models, LPs indicate that visual inputs can significantly contribute to detect MIAs -- AUC>60% is reached in 85.90% of experiments. △ Less

Submitted 10 January, 2025; v1 submitted 29 November, 2024; originally announced November 2024.

Self-Supervised Multi-View Learning for Disentangled Music Audio Representations

Authors: Julia Wilkins, Sivan Ding, Magdalena Fuentes, Juan Pablo Bello

Abstract: Self-supervised learning (SSL) offers a powerful way to learn robust, generalizable representations without labeled data. In music, where labeled data is scarce, existing SSL methods typically use generated supervision and multi-view redundancy to create pretext tasks. However, these approaches often produce entangled representations and lose view-specific information. We propose a novel self-supe… ▽ More Self-supervised learning (SSL) offers a powerful way to learn robust, generalizable representations without labeled data. In music, where labeled data is scarce, existing SSL methods typically use generated supervision and multi-view redundancy to create pretext tasks. However, these approaches often produce entangled representations and lose view-specific information. We propose a novel self-supervised multi-view learning framework for audio designed to incentivize separation between private and shared representation spaces. A case study on audio disentanglement in a controlled setting demonstrates the effectiveness of our method. △ Less

Submitted 4 November, 2024; originally announced November 2024.

Comments: Late Breaking Demo at ISMIR 2024. https://juliawilkins.github.io/marlbymarl/

SONIQUE: Video Background Music Generation Using Unpaired Audio-Visual Data

Authors: Liqian Zhang, Magdalena Fuentes

Abstract: We present SONIQUE, a model for generating background music tailored to video content. Unlike traditional video-to-music generation approaches, which rely heavily on paired audio-visual datasets, SONIQUE leverages unpaired data, combining royalty-free music and independent video sources. By utilizing large language models (LLMs) for video understanding and converting visual descriptions into music… ▽ More We present SONIQUE, a model for generating background music tailored to video content. Unlike traditional video-to-music generation approaches, which rely heavily on paired audio-visual datasets, SONIQUE leverages unpaired data, combining royalty-free music and independent video sources. By utilizing large language models (LLMs) for video understanding and converting visual descriptions into musical tags, alongside a U-Net-based conditional diffusion model, SONIQUE enables customizable music generation. Users can control specific aspects of the music, such as instruments, genres, tempo, and melodies, ensuring the generated output fits their creative vision. SONIQUE is open-source, with a demo available online. △ Less

Submitted 25 February, 2025; v1 submitted 4 October, 2024; originally announced October 2024.

Comments: The paper has been accepted for ICASSP 2025, updating the latest camera-ready version

Efficient and Private Marginal Reconstruction with Local Non-Negativity

Authors: Brett Mullins, Miguel Fuentes, Yingtai Xiao, Daniel Kifer, Cameron Musco, Daniel Sheldon

Abstract: Differential privacy is the dominant standard for formal and quantifiable privacy and has been used in major deployments that impact millions of people. Many differentially private algorithms for query release and synthetic data contain steps that reconstruct answers to queries from answers to other queries that have been measured privately. Reconstruction is an important subproblem for such mecha… ▽ More Differential privacy is the dominant standard for formal and quantifiable privacy and has been used in major deployments that impact millions of people. Many differentially private algorithms for query release and synthetic data contain steps that reconstruct answers to queries from answers to other queries that have been measured privately. Reconstruction is an important subproblem for such mechanisms to economize the privacy budget, minimize error on reconstructed answers, and allow for scalability to high-dimensional datasets. In this paper, we introduce a principled and efficient postprocessing method ReM (Residuals-to-Marginals) for reconstructing answers to marginal queries. Our method builds on recent work on efficient mechanisms for marginal query release, based on making measurements using a residual query basis that admits efficient pseudoinversion, which is an important primitive used in reconstruction. An extension GReM-LNN (Gaussian Residuals-to-Marginals with Local Non-negativity) reconstructs marginals under Gaussian noise satisfying consistency and non-negativity, which often reduces error on reconstructed answers. We demonstrate the utility of ReM and GReM-LNN by applying them to improve existing private query answering mechanisms. △ Less

Submitted 7 December, 2024; v1 submitted 1 October, 2024; originally announced October 2024.

Comments: To appear at NeurIPS 2024

A Critical Assessment of Visual Sound Source Localization Models Including Negative Audio

Authors: Xavier Juanola, Gloria Haro, Magdalena Fuentes

Abstract: The task of Visual Sound Source Localization (VSSL) involves identifying the location of sound sources in visual scenes, integrating audio-visual data for enhanced scene understanding. Despite advancements in state-of-the-art (SOTA) models, we observe three critical flaws: i) The evaluation of the models is mainly focused in sounds produced by objects that are visible in the image, ii) The evaluat… ▽ More The task of Visual Sound Source Localization (VSSL) involves identifying the location of sound sources in visual scenes, integrating audio-visual data for enhanced scene understanding. Despite advancements in state-of-the-art (SOTA) models, we observe three critical flaws: i) The evaluation of the models is mainly focused in sounds produced by objects that are visible in the image, ii) The evaluation often assumes a prior knowledge of the size of the sounding object, and iii) No universal threshold for localization in real-world scenarios is established, as previous approaches only consider positive examples without accounting for both positive and negative cases. In this paper, we introduce a novel test set and metrics designed to complete the current standard evaluation of VSSL models by testing them in scenarios where none of the objects in the image corresponds to the audio input, i.e. a negative audio. We consider three types of negative audio: silence, noise and offscreen. Our analysis reveals that numerous SOTA models fail to appropriately adjust their predictions based on audio input, suggesting that these models may not be leveraging audio information as intended. Additionally, we provide a comprehensive analysis of the range of maximum values in the estimated audio-visual similarity maps, in both positive and negative audio cases, and show that most of the models are not discriminative enough, making them unfit to choose a universal threshold appropriate to perform sound localization without any a priori information of the sounding object, that is, object size and visibility. △ Less

Submitted 10 January, 2025; v1 submitted 1 October, 2024; originally announced October 2024.

Comments: Accepted in ICASSP 2025

Nob-MIAs: Non-biased Membership Inference Attacks Assessment on Large Language Models with Ex-Post Dataset Construction

Authors: Cédric Eichler, Nathan Champeil, Nicolas Anciaux, Alexandra Bensamoun, Heber Hwang Arcolezi, José Maria De Fuentes

Abstract: The rise of Large Language Models (LLMs) has triggered legal and ethical concerns, especially regarding the unauthorized use of copyrighted materials in their training datasets. This has led to lawsuits against tech companies accused of using protected content without permission. Membership Inference Attacks (MIAs) aim to detect whether specific documents were used in a given LLM pretraining, but… ▽ More The rise of Large Language Models (LLMs) has triggered legal and ethical concerns, especially regarding the unauthorized use of copyrighted materials in their training datasets. This has led to lawsuits against tech companies accused of using protected content without permission. Membership Inference Attacks (MIAs) aim to detect whether specific documents were used in a given LLM pretraining, but their effectiveness is undermined by biases such as time-shifts and n-gram overlaps. This paper addresses the evaluation of MIAs on LLMs with partially inferable training sets, under the ex-post hypothesis, which acknowledges inherent distributional biases between members and non-members datasets. We propose and validate algorithms to create ``non-biased'' and ``non-classifiable'' datasets for fairer MIA assessment. Experiments using the Gutenberg dataset on OpenLamma and Pythia show that neutralizing known biases alone is insufficient. Our methods produce non-biased ex-post datasets with AUC-ROC scores comparable to those previously obtained on genuinely random datasets, validating our approach. Globally, MIAs yield results close to random, with only one being effective on both random and our datasets, but its performance decreases when bias is removed. △ Less

Submitted 26 September, 2024; v1 submitted 12 August, 2024; originally announced August 2024.

Joint Selection: Adaptively Incorporating Public Information for Private Synthetic Data

Authors: Miguel Fuentes, Brett Mullins, Ryan McKenna, Gerome Miklau, Daniel Sheldon

Abstract: Mechanisms for generating differentially private synthetic data based on marginals and graphical models have been successful in a wide range of settings. However, one limitation of these methods is their inability to incorporate public data. Initializing a data generating model by pre-training on public data has shown to improve the quality of synthetic data, but this technique is not applicable w… ▽ More Mechanisms for generating differentially private synthetic data based on marginals and graphical models have been successful in a wide range of settings. However, one limitation of these methods is their inability to incorporate public data. Initializing a data generating model by pre-training on public data has shown to improve the quality of synthetic data, but this technique is not applicable when model structure is not determined a priori. We develop the mechanism jam-pgm, which expands the adaptive measurements framework to jointly select between measuring public data and private data. This technique allows for public data to be included in a graphical-model-based mechanism. We show that jam-pgm is able to outperform both publicly assisted and non publicly assisted synthetic data generation mechanisms even when the public data distribution is biased. △ Less

Submitted 12 March, 2024; originally announced March 2024.

Two vs. Four-Channel Sound Event Localization and Detection

Authors: Julia Wilkins, Magdalena Fuentes, Luca Bondi, Shabnam Ghaffarzadegan, Ali Abavisani, Juan Pablo Bello

Abstract: Sound event localization and detection (SELD) systems estimate both the direction-of-arrival (DOA) and class of sound sources over time. In the DCASE 2022 SELD Challenge (Task 3), models are designed to operate in a 4-channel setting. While beneficial to further the development of SELD systems using a multichannel recording setup such as first-order Ambisonics (FOA), most consumer electronics devi… ▽ More Sound event localization and detection (SELD) systems estimate both the direction-of-arrival (DOA) and class of sound sources over time. In the DCASE 2022 SELD Challenge (Task 3), models are designed to operate in a 4-channel setting. While beneficial to further the development of SELD systems using a multichannel recording setup such as first-order Ambisonics (FOA), most consumer electronics devices rarely are able to record using more than two channels. For this reason, in this work we investigate the performance of the DCASE 2022 SELD baseline model using three audio input representations: FOA, binaural, and stereo. We perform a novel comparative analysis illustrating the effect of these audio input representations on SELD performance. Crucially, we show that binaural and stereo (i.e. 2-channel) audio-based SELD models are still able to localize and detect sound sources laterally quite well, despite overall performance degrading as less audio information is provided. Further, we segment our analysis by scenes containing varying degrees of sound source polyphony to better understand the effect of audio input representation on localization and detection performance as scene conditions become increasingly complex. △ Less

Submitted 23 September, 2023; originally announced September 2023.

Sound Source Distance Estimation in Diverse and Dynamic Acoustic Conditions

Authors: Saksham Singh Kushwaha, Iran R. Roman, Magdalena Fuentes, Juan Pablo Bello

Abstract: Localizing a moving sound source in the real world involves determining its direction-of-arrival (DOA) and distance relative to a microphone. Advancements in DOA estimation have been facilitated by data-driven methods optimized with large open-source datasets with microphone array recordings in diverse environments. In contrast, estimating a sound source's distance remains understudied. Existing a… ▽ More Localizing a moving sound source in the real world involves determining its direction-of-arrival (DOA) and distance relative to a microphone. Advancements in DOA estimation have been facilitated by data-driven methods optimized with large open-source datasets with microphone array recordings in diverse environments. In contrast, estimating a sound source's distance remains understudied. Existing approaches assume recordings by non-coincident microphones to use methods that are susceptible to differences in room reverberation. We present a CRNN able to estimate the distance of moving sound sources across multiple datasets featuring diverse rooms, outperforming a recently-published approach. We also characterize our model's performance as a function of sound source distance and different training losses. This analysis reveals optimal training using a loss that weighs model errors as an inverse function of the sound source true distance. Our study is the first to demonstrate that sound source distance estimation can be performed across diverse acoustic conditions using deep learning. △ Less

Submitted 17 September, 2023; originally announced September 2023.

Comments: Accepted in WASPAA 2023

Bridging High-Quality Audio and Video via Language for Sound Effects Retrieval from Visual Queries

Authors: Julia Wilkins, Justin Salamon, Magdalena Fuentes, Juan Pablo Bello, Oriol Nieto

Abstract: Finding the right sound effects (SFX) to match moments in a video is a difficult and time-consuming task, and relies heavily on the quality and completeness of text metadata. Retrieving high-quality (HQ) SFX using a video frame directly as the query is an attractive alternative, removing the reliance on text metadata and providing a low barrier to entry for non-experts. Due to the lack of HQ audio… ▽ More Finding the right sound effects (SFX) to match moments in a video is a difficult and time-consuming task, and relies heavily on the quality and completeness of text metadata. Retrieving high-quality (HQ) SFX using a video frame directly as the query is an attractive alternative, removing the reliance on text metadata and providing a low barrier to entry for non-experts. Due to the lack of HQ audio-visual training data, previous work on audio-visual retrieval relies on YouTube (in-the-wild) videos of varied quality for training, where the audio is often noisy and the video of amateur quality. As such it is unclear whether these systems would generalize to the task of matching HQ audio to production-quality video. To address this, we propose a multimodal framework for recommending HQ SFX given a video frame by (1) leveraging large language models and foundational vision-language models to bridge HQ audio and video to create audio-visual pairs, resulting in a highly scalable automatic audio-visual data curation pipeline; and (2) using pre-trained audio and visual encoders to train a contrastive learning-based retrieval system. We show that our system, trained using our automatic data curation pipeline, significantly outperforms baselines trained on in-the-wild data on the task of HQ SFX retrieval for video. Furthermore, while the baselines fail to generalize to this task, our system generalizes well from clean to in-the-wild data, outperforming the baselines on a dataset of YouTube videos despite only being trained on the HQ audio-visual pairs. A user study confirms that people prefer SFX retrieved by our system over the baseline 67% of the time both for HQ and in-the-wild data. Finally, we present ablations to determine the impact of model and data pipeline design choices on downstream retrieval performance. Please visit our project website to listen to and view our SFX retrieval results. △ Less

Submitted 17 August, 2023; originally announced August 2023.

Comments: WASPAA 2023. Project page: https://juliawilkins.github.io/sound-effects-retrieval-from-video/. 4 pages, 2 figures, 2 tables

A Multimodal Prototypical Approach for Unsupervised Sound Classification

Authors: Saksham Singh Kushwaha, Magdalena Fuentes

Abstract: In the context of environmental sound classification, the adaptability of systems is key: which sound classes are interesting depends on the context and the user's needs. Recent advances in text-to-audio retrieval allow for zero-shot audio classification, but performance compared to supervised models remains limited. This work proposes a multimodal prototypical approach that exploits local audio-t… ▽ More In the context of environmental sound classification, the adaptability of systems is key: which sound classes are interesting depends on the context and the user's needs. Recent advances in text-to-audio retrieval allow for zero-shot audio classification, but performance compared to supervised models remains limited. This work proposes a multimodal prototypical approach that exploits local audio-text embeddings to provide more relevant answers to audio queries, augmenting the adaptability of sound detection in the wild. We do this by first using text to query a nearby community of audio embeddings that best characterize each query sound, and select the group's centroids as our prototypes. Second, we compare unseen audio to these prototypes for classification. We perform multiple ablation studies to understand the impact of the embedding models and prompts. Our unsupervised approach improves upon the zero-shot state-of-the-art in three sound recognition benchmarks by an average of 12%. △ Less

Submitted 17 August, 2023; v1 submitted 21 June, 2023; originally announced June 2023.

Comments: Accepted to INTERSPEECH 2023

Adapting Meter Tracking Models to Latin American Music

Authors: Lucas S. Maia, Martín Rocamora, Luiz W. P. Biscainho, Magdalena Fuentes

Abstract: Beat and downbeat tracking models have improved significantly in recent years with the introduction of deep learning methods. However, despite these improvements, several challenges remain. Particularly, the adaptation of available models to underrepresented music traditions in MIR is usually synonymous with collecting and annotating large amounts of data, which is impractical and time-consuming.… ▽ More Beat and downbeat tracking models have improved significantly in recent years with the introduction of deep learning methods. However, despite these improvements, several challenges remain. Particularly, the adaptation of available models to underrepresented music traditions in MIR is usually synonymous with collecting and annotating large amounts of data, which is impractical and time-consuming. Transfer learning, data augmentation, and fine-tuning techniques have been used quite successfully in related tasks and are known to alleviate this bottleneck. Furthermore, when studying these music traditions, models are not required to generalize to multiple mainstream music genres but to perform well in more constrained, homogeneous conditions. In this work, we investigate simple yet effective strategies to adapt beat and downbeat tracking models to two different Latin American music traditions and analyze the feasibility of these adaptations in real-world applications concerning the data and computational requirements. Contrary to common belief, our findings show it is possible to achieve good performance by spending just a few minutes annotating a portion of the data and training a model in a standard CPU machine, with the precise amount of resources needed depending on the task and the complexity of the dataset. △ Less

Submitted 14 April, 2023; originally announced April 2023.

Comments: Accepted at ISMIR 2022. This version was made after a bug fix in the code, which lead to minor modifications in the results (updated in Figure 1 and Table 1). The paper's conclusions remain unchanged

Tempo vs. Pitch: understanding self-supervised tempo estimation

Authors: Giovana Morais, Matthew E. P. Davies, Marcelo Queiroz, Magdalena Fuentes

Abstract: Self-supervision methods learn representations by solving pretext tasks that do not require human-generated labels, alleviating the need for time-consuming annotations. These methods have been applied in computer vision, natural language processing, environmental sound analysis, and recently in music information retrieval, e.g. for pitch estimation. Particularly in the context of music, there are… ▽ More Self-supervision methods learn representations by solving pretext tasks that do not require human-generated labels, alleviating the need for time-consuming annotations. These methods have been applied in computer vision, natural language processing, environmental sound analysis, and recently in music information retrieval, e.g. for pitch estimation. Particularly in the context of music, there are few insights about the fragility of these models regarding different distributions of data, and how they could be mitigated. In this paper, we explore these questions by dissecting a self-supervised model for pitch estimation adapted for tempo estimation via rigorous experimentation with synthetic data. Specifically, we study the relationship between the input representation and data distribution for self-supervised tempo estimation. △ Less

Submitted 13 April, 2023; originally announced April 2023.

Comments: 5 pages, 3 figures, published on 2023 IEEE International Conference on Acoustics, Speech, and Signal Processing

FlowGrad: Using Motion for Visual Sound Source Localization

Authors: Rajsuryan Singh, Pablo Zinemanas, Xavier Serra, Juan Pablo Bello, Magdalena Fuentes

Abstract: Most recent work in visual sound source localization relies on semantic audio-visual representations learned in a self-supervised manner, and by design excludes temporal information present in videos. While it proves to be effective for widely used benchmark datasets, the method falls short for challenging scenarios like urban traffic. This work introduces temporal context into the state-of-the-ar… ▽ More Most recent work in visual sound source localization relies on semantic audio-visual representations learned in a self-supervised manner, and by design excludes temporal information present in videos. While it proves to be effective for widely used benchmark datasets, the method falls short for challenging scenarios like urban traffic. This work introduces temporal context into the state-of-the-art methods for sound source localization in urban scenes using optical flow as a means to encode motion information. An analysis of the strengths and weaknesses of our methods helps us better understand the problem of visual sound source localization and sheds light on open challenges for audio-visual scene understanding. △ Less

Submitted 14 April, 2023; v1 submitted 15 November, 2022; originally announced November 2022.

Comments: Accepted in ICASSP 2023

How to Listen? Rethinking Visual Sound Localization

Authors: Ho-Hsiang Wu, Magdalena Fuentes, Prem Seetharaman, Juan Pablo Bello

Abstract: Localizing visual sounds consists on locating the position of objects that emit sound within an image. It is a growing research area with potential applications in monitoring natural and urban environments, such as wildlife migration and urban traffic. Previous works are usually evaluated with datasets having mostly a single dominant visible object, and proposed models usually require the introduc… ▽ More Localizing visual sounds consists on locating the position of objects that emit sound within an image. It is a growing research area with potential applications in monitoring natural and urban environments, such as wildlife migration and urban traffic. Previous works are usually evaluated with datasets having mostly a single dominant visible object, and proposed models usually require the introduction of localization modules during training or dedicated sampling strategies, but it remains unclear how these design choices play a role in the adaptability of these methods in more challenging scenarios. In this work, we analyze various model choices for visual sound localization and discuss how their different components affect the model's performance, namely the encoders' architecture, the loss function and the localization strategy. Furthermore, we study the interaction between these decisions, the model performance, and the data, by digging into different evaluation datasets spanning different difficulties and characteristics, and discuss the implications of such decisions in the context of real-world applications. Our code and model weights are open-sourced and made available for further applications. △ Less

Submitted 11 April, 2022; originally announced April 2022.

Comments: Submitted to INTERSPEECH 2022

A Study on Robustness to Perturbations for Representations of Environmental Sound

Authors: Sangeeta Srivastava, Ho-Hsiang Wu, Joao Rulff, Magdalena Fuentes, Mark Cartwright, Claudio Silva, Anish Arora, Juan Pablo Bello

Abstract: Audio applications involving environmental sound analysis increasingly use general-purpose audio representations, also known as embeddings, for transfer learning. Recently, Holistic Evaluation of Audio Representations (HEAR) evaluated twenty-nine embedding models on nineteen diverse tasks. However, the evaluation's effectiveness depends on the variation already captured within a given dataset. The… ▽ More Audio applications involving environmental sound analysis increasingly use general-purpose audio representations, also known as embeddings, for transfer learning. Recently, Holistic Evaluation of Audio Representations (HEAR) evaluated twenty-nine embedding models on nineteen diverse tasks. However, the evaluation's effectiveness depends on the variation already captured within a given dataset. Therefore, for a given data domain, it is unclear how the representations would be affected by the variations caused by myriad microphones' range and acoustic conditions -- commonly known as channel effects. We aim to extend HEAR to evaluate invariance to channel effects in this work. To accomplish this, we imitate channel effects by injecting perturbations to the audio signal and measure the shift in the new (perturbed) embeddings with three distance measures, making the evaluation domain-dependent but not task-dependent. Combined with the downstream performance, it helps us make a more informed prediction of how robust the embeddings are to the channel effects. We evaluate two embeddings -- YAMNet, and OpenL3 on monophonic (UrbanSound8K) and polyphonic (SONYC-UST) urban datasets. We show that one distance measure does not suffice in such task-independent evaluation. Although Fréchet Audio Distance (FAD) correlates with the trend of the performance drop in the downstream task most accurately, we show that we need to study FAD in conjunction with the other distances to get a clear understanding of the overall effect of the perturbation. In terms of the embedding performance, we find OpenL3 to be more robust than YAMNet, which aligns with the HEAR evaluation. △ Less

Submitted 6 July, 2022; v1 submitted 19 March, 2022; originally announced March 2022.

Comments: Accepted in EUSIPCO 2022

Soundata: A Python library for reproducible use of audio datasets

Authors: Magdalena Fuentes, Justin Salamon, Pablo Zinemanas, Martín Rocamora, Genís Paja, Irán R. Román, Marius Miron, Xavier Serra, Juan Pablo Bello

Abstract: Soundata is a Python library for loading and working with audio datasets in a standardized way, removing the need for writing custom loaders in every project, and improving reproducibility by providing tools to validate data against a canonical version. It speeds up research pipelines by allowing users to quickly download a dataset, load it into memory in a standardized and reproducible way, valid… ▽ More Soundata is a Python library for loading and working with audio datasets in a standardized way, removing the need for writing custom loaders in every project, and improving reproducibility by providing tools to validate data against a canonical version. It speeds up research pipelines by allowing users to quickly download a dataset, load it into memory in a standardized and reproducible way, validate that the dataset is complete and correct, and more. Soundata is based and inspired on mirdata and design to complement mirdata by working with environmental sound, bioacoustic and speech datasets, among others. Soundata was created to be easy to use, easy to contribute to, and to increase reproducibility and standardize usage of sound datasets in a flexible way. △ Less

Submitted 4 October, 2021; v1 submitted 26 September, 2021; originally announced September 2021.

Exploring modality-agnostic representations for music classification

Authors: Ho-Hsiang Wu, Magdalena Fuentes, Juan P. Bello

Abstract: Music information is often conveyed or recorded across multiple data modalities including but not limited to audio, images, text and scores. However, music information retrieval research has almost exclusively focused on single modality recognition, requiring development of separate models for each modality. Some multi-modal works require multiple coexisting modalities given to the model as inputs… ▽ More Music information is often conveyed or recorded across multiple data modalities including but not limited to audio, images, text and scores. However, music information retrieval research has almost exclusively focused on single modality recognition, requiring development of separate models for each modality. Some multi-modal works require multiple coexisting modalities given to the model as inputs, constraining the use of these models to the few cases where data from all modalities are available. To the best of our knowledge, no existing model has the ability to take inputs from varying modalities, e.g. images or sounds, and classify them into unified music categories. We explore the use of cross-modal retrieval as a pretext task to learn modality-agnostic representations, which can then be used as inputs to classifiers that are independent of modality. We select instrument classification as an example task for our study as both visual and audio components provide relevant semantic information. We train music instrument classifiers that can take both images or sounds as input, and perform comparably to sound-only or image-only classifiers. Furthermore, we explore the case when there is limited labeled data for a given modality, and the impact in performance by using labeled data from other modalities. We are able to achieve almost 70% of best performing system in a zero-shot setting. We provide a detailed analysis of experimental results to understand the potential and limitations of the approach, and discuss future steps towards modality-agnostic classifiers. △ Less

Submitted 2 June, 2021; originally announced June 2021.

SONYC-UST-V2: An Urban Sound Tagging Dataset with Spatiotemporal Context

Authors: Mark Cartwright, Jason Cramer, Ana Elisa Mendez Mendez, Yu Wang, Ho-Hsiang Wu, Vincent Lostanlen, Magdalena Fuentes, Graham Dove, Charlie Mydlarz, Justin Salamon, Oded Nov, Juan Pablo Bello

Abstract: We present SONYC-UST-V2, a dataset for urban sound tagging with spatiotemporal information. This dataset is aimed for the development and evaluation of machine listening systems for real-world urban noise monitoring. While datasets of urban recordings are available, this dataset provides the opportunity to investigate how spatiotemporal metadata can aid in the prediction of urban sound tags. SONYC… ▽ More We present SONYC-UST-V2, a dataset for urban sound tagging with spatiotemporal information. This dataset is aimed for the development and evaluation of machine listening systems for real-world urban noise monitoring. While datasets of urban recordings are available, this dataset provides the opportunity to investigate how spatiotemporal metadata can aid in the prediction of urban sound tags. SONYC-UST-V2 consists of 18510 audio recordings from the "Sounds of New York City" (SONYC) acoustic sensor network, including the timestamp of audio acquisition and location of the sensor. The dataset contains annotations by volunteers from the Zooniverse citizen science platform, as well as a two-stage verification with our team. In this article, we describe our data collection procedure and propose evaluation metrics for multilabel classification of urban sound tags. We report the results of a simple baseline model that exploits spatiotemporal information. △ Less

Submitted 10 September, 2020; originally announced September 2020.

Pioneering Studies on LTE eMBMS: Towards 5G Point-to-Multipoint Transmissions

Authors: Hongzhi Chen, De Mi, Manuel Fuentes, David Vargas, Eduardo Garro, Jose Luis Carcel, Belkacem Mouhouche, Pei Xiao, Rahim Tafazolli

Abstract: The first 5G (5th generation wireless systems) New Radio Release-15 was recently completed. However, the specification only considers the use of unicast technologies and the extension to point-to-multipoint (PTM) scenarios is not yet considered. To this end, we first present in this work a technical overview of the state-of-the-art LTE (Long Term Evolution) PTM technology, i.e., eMBMS (evolved Mul… ▽ More The first 5G (5th generation wireless systems) New Radio Release-15 was recently completed. However, the specification only considers the use of unicast technologies and the extension to point-to-multipoint (PTM) scenarios is not yet considered. To this end, we first present in this work a technical overview of the state-of-the-art LTE (Long Term Evolution) PTM technology, i.e., eMBMS (evolved Multimedia Broadcast Multicast Services), and investigate the physical layer performance via link-level simulations. Then based on the simulation analysis, we discuss potential improvements for the two current eMBMS solutions, i.e., MBSFN (MBMS over Single Frequency Networks) and SC-PTM (Single-Cell PTM). This work explicitly focus on equipping the current eMBMS solutions with 5G candidate techniques, e.g., multiple antennas and millimeter wave, and its potentials to meet the requirements of next generation PTM transmissions. △ Less

Submitted 29 November, 2019; originally announced January 2020.

Comments: SAM 2018, 5 pages, 4 figs

On the Performance of PDCCH in LTE and 5G New Radio

Authors: Hongzhi Chen, De Mi, Manuel Fuentes, Eduardo Garro, Jose Luis Carcel, Belkacem Mouhouche, Pei Xiao, Rahim Tafazolli

Abstract: 5G New Radio (NR) Release 15 has been specified in June 2018. It introduces numerous changes and potential improvements for physical layer data transmissions, although only point-to-point (PTP) communications are considered. In order to use physical data channels such as the Physical Downlink Shared Channel (PDSCH), it is essential to guarantee a successful transmission of control information via… ▽ More 5G New Radio (NR) Release 15 has been specified in June 2018. It introduces numerous changes and potential improvements for physical layer data transmissions, although only point-to-point (PTP) communications are considered. In order to use physical data channels such as the Physical Downlink Shared Channel (PDSCH), it is essential to guarantee a successful transmission of control information via the Physical Downlink Control Channel (PDCCH). Taking into account these two aspects, in this paper, we first analyze the PDCCH processing chain in NR PTP as well as in the state-of-the-art Long Term Evolution (LTE) point-to-multipoint (PTM) solution, i.e., evolved Multimedia Broadcast Multicast Service (eMBMS). Then, via link level simulations, we compare the performance of the two technologies, observing the Bit/Block Error Rate (BER/BLER) for various scenarios. The objective is to identify the performance gap brought by physical layer changes in NR PDCCH as well as provide insightful guidelines on the control channel configuration towards NR PTM scenarios. △ Less

Submitted 29 November, 2019; originally announced January 2020.

Comments: Globecomm 2018 workshop, 6 pages, 7 figs

A Fuzzy Control System for Inductive Video Games

Authors: Carlos Lara-Alvarez, Hugo Mitre-Hernandez, Juan Flores, Maria Fuentes

Abstract: It has been shown that the emotional state of students has an important relationship with learning; for instance, engaged concentration is positively correlated with learning. This paper proposes the Inductive Control (IC) for educational games. Unlike conventional approaches that only modify the game level, the proposed technique also induces emotions in the player for supporting the learning pro… ▽ More It has been shown that the emotional state of students has an important relationship with learning; for instance, engaged concentration is positively correlated with learning. This paper proposes the Inductive Control (IC) for educational games. Unlike conventional approaches that only modify the game level, the proposed technique also induces emotions in the player for supporting the learning process. This paper explores a fuzzy system that analyzes the players' performance and their emotional state for controlling the level and aesthetic content of an educational video game. The emotional state of the player is recognized through voice analysis. A total of 20 subjects played a video game designed to practice basic math skills; for each trial, a student plays two times in a row the same game but each time the game was controlled by one of the two approaches ---Dynamic Difficulty Adjustment (DDA) and IC, the playing order was assigned randomly. Results show that when the proposed approach is used the participants changed faster from Unpleasant--low to pleasant or high emotions, and reached softly and kept in the flow zone. These experiments demonstrate that the inductive control technique improves the learning effectiveness through detection and stimulation of positive emotions. △ Less

Submitted 15 April, 2018; v1 submitted 4 September, 2017; originally announced September 2017.

Comments: It needs to be reviewed

Does network complexity help organize Babel's library?

Authors: Juan Pablo Cárdenas, Iván González, Gerardo Vidal, Miguel Fuentes

Abstract: In this work, we study properties of texts from the perspective of complex network theory. Words in given texts are linked by co-occurrence and transformed into networks, and we observe that these display topological properties common to other complex systems. However, there are some properties that seem to be exclusive to texts; many of these properties depend on the frequency of words in the tex… ▽ More In this work, we study properties of texts from the perspective of complex network theory. Words in given texts are linked by co-occurrence and transformed into networks, and we observe that these display topological properties common to other complex systems. However, there are some properties that seem to be exclusive to texts; many of these properties depend on the frequency of words in the text, while others seem to be strictly determined by the grammar. Precisely, these properties allow for a categorization of texts as either with a sense and others encoded or senseless. △ Less

Submitted 16 October, 2015; v1 submitted 23 September, 2014; originally announced September 2014.