-
Steering dense music retrieval with open-vocabulary concept discovery
Authors:
Julien Guinot,
Alain Riou,
Elio Quinton,
György Fazekas
Abstract:
Controllable music retrieval lets users find music that is, for example, more ambient, less distorted, or without guitar while preserving the other semantic content of an original seed query. Sparse autoencoders (SAEs) are a promising interface for this kind of concept-level control, but a key problem remains: given a free-form text concept, which sparse features should be edited? In shared multim…
▽ More
Controllable music retrieval lets users find music that is, for example, more ambient, less distorted, or without guitar while preserving the other semantic content of an original seed query. Sparse autoencoders (SAEs) are a promising interface for this kind of concept-level control, but a key problem remains: given a free-form text concept, which sparse features should be edited? In shared multimodal embedding spaces, standard attribution methods often select neurons that match the concept's wording but not the audio examples that express it. This leads to weak or unstable edits: relevant features are missed when concepts are distributed across neurons, while others are selected due to text alignment rather than audio-side structure.
We address this with a lightweight, training-free method that recovers a sparse set of audio features whose decoded representation reconstructs the target concept while remaining consistent with audio-space geometry. This reframes concept attribution as a sparse inversion problem rather than a text-side neuron-ranking heuristic. The method requires neither paired audio-text supervision nor SAE retraining. We evaluate this approach in steerable music retrieval and show that the recovered supports align more closely with concept-bearing audio examples and achieve a stronger trade-off between edit strength and preservation than alignment baselines, enabling more precise concept amplification and suppression with reduced drift on preservation metrics.
△ Less
Submitted 9 August, 2026;
originally announced August 2026.
-
On the Geometry of Music Bandwidth Extension in Latent Spaces of Audio Codecs
Authors:
Hendrik Vincent Koops,
Hao Hao Tan,
Elio Quinton
Abstract:
Recent audio restoration increasingly relies on large-scale conditional latent generative modeling, including diffusion, Schrodinger Bridges, and Flow Matching variants, to invert degradations such as bandwidth limitation or noise. We present an analysis of the performance of various state-of-the-art methods compared to simple arithmetic transformations in the latent spaces of multiple neural code…
▽ More
Recent audio restoration increasingly relies on large-scale conditional latent generative modeling, including diffusion, Schrodinger Bridges, and Flow Matching variants, to invert degradations such as bandwidth limitation or noise. We present an analysis of the performance of various state-of-the-art methods compared to simple arithmetic transformations in the latent spaces of multiple neural codecs for musical bandwidth extension. We show that estimating a single transport vector between the clean and degraded latent centroids on a reference set, and adding it to degraded latents, can yield restoration performance competitive with large diffusion models. This suggests, first, that some neural codec latent spaces exhibit structure aligned with audio bandwidth; and second, that in such cases complex conditional models may offer only limited gains over a simple vector addition. We argue that these findings reveal an interesting avenue for future research whereby models could take advantage of the latent space structure in order to offer greater training and parameter efficiency, and overall better performance. Additionally, we propose to consider this simple arithmetic transformation as a baseline for music bandwidth extension research, as it allows an assessment of the contribution of learnable parameters towards restoration performance.
△ Less
Submitted 4 August, 2026;
originally announced August 2026.
-
Single-step Controllable Music Bandwidth Extension With Flow Matching
Authors:
Carlos Hernandez-Olivan,
Hendrik Vincent Koops,
Hao Hao Tan,
Elio Quinton
Abstract:
Audio restoration consists in inverting degradations of a digital audio signal to recover what would have been the pristine quality signal before the degradation occurred. This is valuable in contexts such as archives of music recordings, particularly those of precious historical value, for which a clean version may have been lost or simply does not exist. Recent work applied generative models to…
▽ More
Audio restoration consists in inverting degradations of a digital audio signal to recover what would have been the pristine quality signal before the degradation occurred. This is valuable in contexts such as archives of music recordings, particularly those of precious historical value, for which a clean version may have been lost or simply does not exist. Recent work applied generative models to audio restoration, showing promising improvement over previous methods, and opening the door to the ability to perform restoration operations that were not possible before. However, making these models finely controllable remains a challenge. In this paper, we propose an extension of FLowHigh and introduce the Dynamic Spectral Contour (DSC) as a control signal for bandwidth extension via classifier-free guidance. Our experiments show competitive model performance, and indicate that DSC is a promising feature to support fine-grained conditioning.
△ Less
Submitted 20 January, 2026;
originally announced January 2026.
-
Not All Deepfakes Are Created Equal: Triaging Audio Forgeries for Robust Deepfake Singer Identification
Authors:
Davide Salvi,
Hendrik Vincent Koops,
Elio Quinton
Abstract:
The proliferation of highly realistic singing voice deepfakes presents a significant challenge to protecting artist likeness and content authenticity. Automatic singer identification in vocal deepfakes is a promising avenue for artists and rights holders to defend against unauthorized use of their voice, but remains an open research problem. Based on the premise that the most harmful deepfakes are…
▽ More
The proliferation of highly realistic singing voice deepfakes presents a significant challenge to protecting artist likeness and content authenticity. Automatic singer identification in vocal deepfakes is a promising avenue for artists and rights holders to defend against unauthorized use of their voice, but remains an open research problem. Based on the premise that the most harmful deepfakes are those of the highest quality, we introduce a two-stage pipeline to identify a singer's vocal likeness. It first employs a discriminator model to filter out low-quality forgeries that fail to accurately reproduce vocal likeness. A subsequent model, trained exclusively on authentic recordings, identifies the singer in the remaining high-quality deepfakes and authentic audio. Experiments show that this system consistently outperforms existing baselines on both authentic and synthetic content.
△ Less
Submitted 18 November, 2025; v1 submitted 20 October, 2025;
originally announced October 2025.
-
GD-Retriever: Controllable Generative Text-Music Retrieval with Diffusion Models
Authors:
Julien Guinot,
Elio Quinton,
György Fazekas
Abstract:
Multimodal contrastive models have achieved strong performance in text-audio retrieval and zero-shot settings, but improving joint embedding spaces remains an active research area. Less attention has been given to making these systems controllable and interactive for users. In text-music retrieval, the ambiguity of freeform language creates a many-to-many mapping, often resulting in inflexible or…
▽ More
Multimodal contrastive models have achieved strong performance in text-audio retrieval and zero-shot settings, but improving joint embedding spaces remains an active research area. Less attention has been given to making these systems controllable and interactive for users. In text-music retrieval, the ambiguity of freeform language creates a many-to-many mapping, often resulting in inflexible or unsatisfying results.
We introduce Generative Diffusion Retriever (GDR), a novel framework that leverages diffusion models to generate queries in a retrieval-optimized latent space. This enables controllability through generative tools such as negative prompting and denoising diffusion implicit models (DDIM) inversion, opening a new direction in retrieval control. GDR improves retrieval performance over contrastive teacher models and supports retrieval in audio-only latent spaces using non-jointly trained encoders. Finally, we demonstrate that GDR enables effective post-hoc manipulation of retrieval behavior, enhancing interactive control for text-music retrieval tasks.
△ Less
Submitted 24 June, 2025; v1 submitted 21 June, 2025;
originally announced June 2025.
-
SLAP: Siamese Language-Audio Pretraining Without Negative Samples for Music Understanding
Authors:
Julien Guinot,
Alain Riou,
Elio Quinton,
György Fazekas
Abstract:
Joint embedding spaces have significantly advanced music understanding and generation by linking text and audio through multimodal contrastive learning. However, these approaches face large memory requirement limitations due to relying on large batch sizes to effectively utilize negative samples. Further, multimodal joint embedding spaces suffer from a modality gap wherein embeddings from differen…
▽ More
Joint embedding spaces have significantly advanced music understanding and generation by linking text and audio through multimodal contrastive learning. However, these approaches face large memory requirement limitations due to relying on large batch sizes to effectively utilize negative samples. Further, multimodal joint embedding spaces suffer from a modality gap wherein embeddings from different modalities lie in different manifolds of the embedding space. To address these challenges, we propose Siamese Language-Audio Pretraining (SLAP), a novel multimodal pretraining framework that allows learning powerful representations without negative samples. SLAP adapts the Bootstrap Your Own Latent (BYOL) paradigm for multimodal audio-text training, promoting scalability in training multimodal embedding spaces.
We illustrate the ability of our model to learn meaningful relationships between music and text -- specifically, we show that SLAP outperforms CLAP on tasks such as text-music retrieval and zero-shot classification. We also observe competitive downstream performance on several MIR tasks, including with larger or supervised models (genre and instrument classification, auto-tagging). Additionally, our approach has attractive properties, such as a quantifiably reduced modality gap and improved robustness to batch size variations on retrieval performance. Finally, its novel formulation unlocks large-scale training on a single GPU through gradient accumulation.
△ Less
Submitted 21 June, 2025;
originally announced June 2025.
-
Towards a Unified Representation Evaluation Framework Beyond Downstream Tasks
Authors:
Christos Plachouras,
Julien Guinot,
George Fazekas,
Elio Quinton,
Emmanouil Benetos,
Johan Pauwels
Abstract:
Downstream probing has been the dominant method for evaluating model representations, an important process given the increasing prominence of self-supervised learning and foundation models. However, downstream probing primarily assesses the availability of task-relevant information in the model's latent space, overlooking attributes such as equivariance, invariance, and disentanglement, which cont…
▽ More
Downstream probing has been the dominant method for evaluating model representations, an important process given the increasing prominence of self-supervised learning and foundation models. However, downstream probing primarily assesses the availability of task-relevant information in the model's latent space, overlooking attributes such as equivariance, invariance, and disentanglement, which contribute to the interpretability, adaptability, and utility of representations in real-world applications. While some attempts have been made to measure these qualities in representations, no unified evaluation framework with modular, generalizable, and interpretable metrics exists.
In this paper, we argue for the importance of representation evaluation beyond downstream probing. We introduce a standardized protocol to quantify informativeness, equivariance, invariance, and disentanglement of factors of variation in model representations. We use it to evaluate representations from a variety of models in the image and speech domains using different architectures and pretraining approaches on identified controllable factors of variation. We find that representations from models with similar downstream performance can behave substantially differently with regard to these attributes. This hints that the respective mechanisms underlying their downstream performance are functionally different, prompting new research directions to understand and improve representations.
△ Less
Submitted 9 May, 2025;
originally announced May 2025.
-
Leave-One-EquiVariant: Alleviating invariance-related information loss in contrastive music representations
Authors:
Julien Guinot,
Elio Quinton,
György Fazekas
Abstract:
Contrastive learning has proven effective in self-supervised musical representation learning, particularly for Music Information Retrieval (MIR) tasks. However, reliance on augmentation chains for contrastive view generation and the resulting learnt invariances pose challenges when different downstream tasks require sensitivity to certain musical attributes. To address this, we propose the Leave O…
▽ More
Contrastive learning has proven effective in self-supervised musical representation learning, particularly for Music Information Retrieval (MIR) tasks. However, reliance on augmentation chains for contrastive view generation and the resulting learnt invariances pose challenges when different downstream tasks require sensitivity to certain musical attributes. To address this, we propose the Leave One EquiVariant (LOEV) framework, which introduces a flexible, task-adaptive approach compared to previous work by selectively preserving information about specific augmentations, allowing the model to maintain task-relevant equivariances. We demonstrate that LOEV alleviates information loss related to learned invariances, improving performance on augmentation related tasks and retrieval without sacrificing general representation quality. Furthermore, we introduce a variant of LOEV, LOEV++, which builds a disentangled latent space by design in a self-supervised manner, and enables targeted retrieval based on augmentation related attributes.
△ Less
Submitted 25 December, 2024;
originally announced December 2024.
-
Exploring trends in audio mixes and masters: Insights from a dataset analysis
Authors:
Angeliki Mourgela,
Elio Quinton,
Spyridon Bissas,
Joshua D. Reiss,
David Ronan
Abstract:
We present an analysis of a dataset of audio metrics and aesthetic considerations about mixes and masters provided by the web platform MixCheck studio. The platform is designed for educational purposes, primarily targeting amateur music producers, and aimed at analysing their recordings prior to them being released. The analysis focuses on the following data points: integrated loudness, mono compa…
▽ More
We present an analysis of a dataset of audio metrics and aesthetic considerations about mixes and masters provided by the web platform MixCheck studio. The platform is designed for educational purposes, primarily targeting amateur music producers, and aimed at analysing their recordings prior to them being released. The analysis focuses on the following data points: integrated loudness, mono compatibility, presence of clipping and phase issues, compression and tonal profile across 30 user-specified genres. Both mixed (mixes) and mastered audio (masters) are included in the analysis, where mixes refer to the initial combination and balance of individual tracks, and masters refer to the final refined version optimized for distribution. Results show that loudness-related issues along with dynamics issues are the most prevalent, particularly in mastered audio. However mastered audio presents better results in compression than just mixed audio. Additionally, results show that mastered audio has a lower percentage of stereo field and phase issues.
△ Less
Submitted 4 December, 2024;
originally announced December 2024.
-
MuChoMusic: Evaluating Music Understanding in Multimodal Audio-Language Models
Authors:
Benno Weck,
Ilaria Manco,
Emmanouil Benetos,
Elio Quinton,
George Fazekas,
Dmitry Bogdanov
Abstract:
Multimodal models that jointly process audio and language hold great promise in audio understanding and are increasingly being adopted in the music domain. By allowing users to query via text and obtain information about a given audio input, these models have the potential to enable a variety of music understanding tasks via language-based interfaces. However, their evaluation poses considerable c…
▽ More
Multimodal models that jointly process audio and language hold great promise in audio understanding and are increasingly being adopted in the music domain. By allowing users to query via text and obtain information about a given audio input, these models have the potential to enable a variety of music understanding tasks via language-based interfaces. However, their evaluation poses considerable challenges, and it remains unclear how to effectively assess their ability to correctly interpret music-related inputs with current methods. Motivated by this, we introduce MuChoMusic, a benchmark for evaluating music understanding in multimodal language models focused on audio. MuChoMusic comprises 1,187 multiple-choice questions, all validated by human annotators, on 644 music tracks sourced from two publicly available music datasets, and covering a wide variety of genres. Questions in the benchmark are crafted to assess knowledge and reasoning abilities across several dimensions that cover fundamental musical concepts and their relation to cultural and functional contexts. Through the holistic analysis afforded by the benchmark, we evaluate five open-source models and identify several pitfalls, including an over-reliance on the language modality, pointing to a need for better multimodal integration. Data and code are open-sourced.
△ Less
Submitted 2 August, 2024;
originally announced August 2024.
-
Robust Lossy Audio Compression Identification
Authors:
Hendrik Vincent Koops,
Gianluca Micchi,
Elio Quinton
Abstract:
Previous research contributions on blind lossy compression identification report near perfect performance metrics on their test set, across a variety of codecs and bit rates. However, we show that such results can be deceptive and may not accurately represent true ability of the system to tackle the task at hand. In this article, we present an investigation into the robustness and generalisation c…
▽ More
Previous research contributions on blind lossy compression identification report near perfect performance metrics on their test set, across a variety of codecs and bit rates. However, we show that such results can be deceptive and may not accurately represent true ability of the system to tackle the task at hand. In this article, we present an investigation into the robustness and generalisation capability of a lossy audio identification model. Our contributions are as follows. (1) We show the lack of robustness to codec parameter variations of a model equivalent to prior art. In particular, when naively training a lossy compression detection model on a dataset of music recordings processed with a range of codecs and their lossless counterparts, we obtain near perfect performance metrics on the held-out test set, but severely degraded performance on lossy tracks produced with codec parameters not seen in training. (2) We propose and show the effectiveness of an improved training strategy to significantly increase the robustness and generalisation capability of the model beyond codec configurations seen during training. Namely we apply a random mask to the input spectrogram to encourage the model not to rely solely on the training set's codec cutoff frequency.
△ Less
Submitted 31 July, 2024;
originally announced July 2024.
-
The Song Describer Dataset: a Corpus of Audio Captions for Music-and-Language Evaluation
Authors:
Ilaria Manco,
Benno Weck,
SeungHeon Doh,
Minz Won,
Yixiao Zhang,
Dmitry Bogdanov,
Yusong Wu,
Ke Chen,
Philip Tovstogan,
Emmanouil Benetos,
Elio Quinton,
György Fazekas,
Juhan Nam
Abstract:
We introduce the Song Describer dataset (SDD), a new crowdsourced corpus of high-quality audio-caption pairs, designed for the evaluation of music-and-language models. The dataset consists of 1.1k human-written natural language descriptions of 706 music recordings, all publicly accessible and released under Creative Common licenses. To showcase the use of our dataset, we benchmark popular models o…
▽ More
We introduce the Song Describer dataset (SDD), a new crowdsourced corpus of high-quality audio-caption pairs, designed for the evaluation of music-and-language models. The dataset consists of 1.1k human-written natural language descriptions of 706 music recordings, all publicly accessible and released under Creative Common licenses. To showcase the use of our dataset, we benchmark popular models on three key music-and-language tasks (music captioning, text-to-music generation and music-language retrieval). Our experiments highlight the importance of cross-dataset evaluation and offer insights into how researchers can use SDD to gain a broader understanding of model performance.
△ Less
Submitted 22 November, 2023; v1 submitted 16 November, 2023;
originally announced November 2023.
-
Serenade: A Model for Human-in-the-loop Automatic Chord Estimation
Authors:
Hendrik Vincent Koops,
Gianluca Micchi,
Ilaria Manco,
Elio Quinton
Abstract:
Computational harmony analysis is important for MIR tasks such as automatic segmentation, corpus analysis and automatic chord label estimation. However, recent research into the ambiguous nature of musical harmony, causing limited inter-rater agreement, has made apparent that there is a glass ceiling for common metrics such as accuracy. Commonly, these issues are addressed either in the training d…
▽ More
Computational harmony analysis is important for MIR tasks such as automatic segmentation, corpus analysis and automatic chord label estimation. However, recent research into the ambiguous nature of musical harmony, causing limited inter-rater agreement, has made apparent that there is a glass ceiling for common metrics such as accuracy. Commonly, these issues are addressed either in the training data itself by creating majority-rule annotations or during the training phase by learning soft targets. We propose a novel alternative approach in which a human and an autoregressive model together co-create a harmonic annotation for an audio track. After automatically generating harmony predictions, a human sparsely annotates parts with low model confidence and the model then adjusts its predictions following human guidance. We evaluate our model on a dataset of popular music and we show that, with this human-in-the-loop approach, harmonic analysis performance improves over a model-only approach. The human contribution is amplified by the second, constrained prediction of the model.
△ Less
Submitted 17 October, 2023;
originally announced October 2023.
-
Equivariant Self-Supervision for Musical Tempo Estimation
Authors:
Elio Quinton
Abstract:
Self-supervised methods have emerged as a promising avenue for representation learning in the recent years since they alleviate the need for labeled datasets, which are scarce and expensive to acquire. Contrastive methods are a popular choice for self-supervision in the audio domain, and typically provide a learning signal by forcing the model to be invariant to some transformations of the input.…
▽ More
Self-supervised methods have emerged as a promising avenue for representation learning in the recent years since they alleviate the need for labeled datasets, which are scarce and expensive to acquire. Contrastive methods are a popular choice for self-supervision in the audio domain, and typically provide a learning signal by forcing the model to be invariant to some transformations of the input. These methods, however, require measures such as negative sampling or some form of regularisation to be taken to prevent the model from collapsing on trivial solutions. In this work, instead of invariance, we propose to use equivariance as a self-supervision signal to learn audio tempo representations from unlabelled data. We derive a simple loss function that prevents the network from collapsing on a trivial solution during training, without requiring any form of regularisation or negative sampling. Our experiments show that it is possible to learn meaningful representations for tempo estimation by solely relying on equivariant self-supervision, achieving performance comparable with supervised methods on several benchmarks. As an added benefit, our method only requires moderate compute resources and therefore remains accessible to a wide research community.
△ Less
Submitted 3 September, 2022;
originally announced September 2022.
-
Contrastive Audio-Language Learning for Music
Authors:
Ilaria Manco,
Emmanouil Benetos,
Elio Quinton,
György Fazekas
Abstract:
As one of the most intuitive interfaces known to humans, natural language has the potential to mediate many tasks that involve human-computer interaction, especially in application-focused fields like Music Information Retrieval. In this work, we explore cross-modal learning in an attempt to bridge audio and language in the music domain. To this end, we propose MusCALL, a framework for Music Contr…
▽ More
As one of the most intuitive interfaces known to humans, natural language has the potential to mediate many tasks that involve human-computer interaction, especially in application-focused fields like Music Information Retrieval. In this work, we explore cross-modal learning in an attempt to bridge audio and language in the music domain. To this end, we propose MusCALL, a framework for Music Contrastive Audio-Language Learning. Our approach consists of a dual-encoder architecture that learns the alignment between pairs of music audio and descriptive sentences, producing multimodal embeddings that can be used for text-to-audio and audio-to-text retrieval out-of-the-box. Thanks to this property, MusCALL can be transferred to virtually any task that can be cast as text-based retrieval. Our experiments show that our method performs significantly better than the baselines at retrieving audio that matches a textual description and, conversely, text that matches an audio query. We also demonstrate that the multimodal alignment capability of our model can be successfully extended to the zero-shot transfer scenario for genre classification and auto-tagging on two public datasets.
△ Less
Submitted 25 August, 2022;
originally announced August 2022.
-
Learning music audio representations via weak language supervision
Authors:
Ilaria Manco,
Emmanouil Benetos,
Elio Quinton,
Gyorgy Fazekas
Abstract:
Audio representations for music information retrieval are typically learned via supervised learning in a task-specific fashion. Although effective at producing state-of-the-art results, this scheme lacks flexibility with respect to the range of applications a model can have and requires extensively annotated datasets. In this work, we pose the question of whether it may be possible to exploit weak…
▽ More
Audio representations for music information retrieval are typically learned via supervised learning in a task-specific fashion. Although effective at producing state-of-the-art results, this scheme lacks flexibility with respect to the range of applications a model can have and requires extensively annotated datasets. In this work, we pose the question of whether it may be possible to exploit weakly aligned text as the only supervisory signal to learn general-purpose music audio representations. To address this question, we design a multimodal architecture for music and language pre-training (MuLaP) optimised via a set of proxy tasks. Weak supervision is provided in the form of noisy natural language descriptions conveying the overall musical content of the track. After pre-training, we transfer the audio backbone of the model to a set of music audio classification and regression tasks. We demonstrate the usefulness of our approach by comparing the performance of audio representations produced by the same audio backbone with different training strategies and show that our pre-training method consistently achieves comparable or higher scores on all tasks and datasets considered. Our experiments also confirm that MuLaP effectively leverages audio-caption pairs to learn representations that are competitive with audio-only and cross-modal self-supervised methods in the literature.
△ Less
Submitted 17 February, 2022; v1 submitted 8 December, 2021;
originally announced December 2021.
-
MusCaps: Generating Captions for Music Audio
Authors:
Ilaria Manco,
Emmanouil Benetos,
Elio Quinton,
Gyorgy Fazekas
Abstract:
Content-based music information retrieval has seen rapid progress with the adoption of deep learning. Current approaches to high-level music description typically make use of classification models, such as in auto-tagging or genre and mood classification. In this work, we propose to address music description via audio captioning, defined as the task of generating a natural language description of…
▽ More
Content-based music information retrieval has seen rapid progress with the adoption of deep learning. Current approaches to high-level music description typically make use of classification models, such as in auto-tagging or genre and mood classification. In this work, we propose to address music description via audio captioning, defined as the task of generating a natural language description of music audio content in a human-like manner. To this end, we present the first music audio captioning model, MusCaps, consisting of an encoder-decoder with temporal attention. Our method combines convolutional and recurrent neural network architectures to jointly process audio-text inputs through a multimodal encoder and leverages pre-training on audio data to obtain representations that effectively capture and summarise musical features in the input. Evaluation of the generated captions through automatic metrics shows that our method outperforms a baseline designed for non-music audio captioning. Through an ablation study, we unveil that this performance boost can be mainly attributed to pre-training of the audio encoder, while other design choices - modality fusion, decoding strategy and the use of attention - contribute only marginally. Our model represents a shift away from classification-based music description and combines tasks requiring both auditory and linguistic understanding to bridge the semantic gap in music information retrieval.
△ Less
Submitted 24 April, 2021;
originally announced April 2021.