Showing 1–26 of 26 results for author: Moradi, A

A Few Cases Are All You Need: An Empirical Study of Annotation-Efficient LoRA Fine-Tuning of MedSAM3

Authors: Sachin Dudda Nagaraju, Bendik Skarre Abrahamsen, Ashkan Moradi, Mattijs Elschot

Abstract: Medical image segmentation is essential for clinical workflows such as treatment planning and disease assessment. While specialist tools like TotalSegmentator and MRSegmentator achieve strong performance, they require large annotated datasets for training. Medical foundation models offer a promising alternative through large-scale pretraining that reduces the annotation burden for new tasks, but z… ▽ More Medical image segmentation is essential for clinical workflows such as treatment planning and disease assessment. While specialist tools like TotalSegmentator and MRSegmentator achieve strong performance, they require large annotated datasets for training. Medical foundation models offer a promising alternative through large-scale pretraining that reduces the annotation burden for new tasks, but zero-shot performance remains limited. Parameter-efficient adaptation via Low-Rank Adaptation (LoRA) enables efficient specialization with few trainable parameters, but a key question remains: how many expert-annotated cases are needed to achieve clinically useful segmentation performance? We address this by adapting MedSAM3 with LoRA for five abdominal organs (liver, kidneys, spleen, gallbladder, and pancreas) in CT and MRI using only 1, 2, 5, and 10 annotated cases, evaluating on AMOS22 dataset. With just 10 cases, models achieve performance competitive with specialist systems trained on orders of magnitude more data. Notably, this includes reliable gallbladder segmentation (Dice 0.68 CT, 0.59 MRI) where existing tools fail almost completely (Dice 0.0004), while remaining within 5--10% of MRSegmentator for liver, kidneys, and spleen using over 100 times fewer annotations. Furthermore, external validation on the Whole Heart Segmentation dataset shows that the approach extends to cardiac segmentation, a use case beyond the scope of TotalSegmentator (MRI) and MRSegmentator, achieving competitive left ventricle (LV) performance with only 10 annotated cases. Training requires only3--5,hours per organ on a single GPU, approximately 2--3 times faster than nnU-Net. These findings suggest that ten annotated cases are sufficient for clinically useful segmentation, effectively reducing bottlenecks for both image annotation and training time. △ Less

Submitted 19 August, 2026; originally announced August 2026.

SMART: A Machine Learning and Monte Carlo Framework for Rapid Analysis of Stochastic Transistor Aging and Process Variation in Digital Circuits

Authors: Arash Esshaghi, Siavash Es'haghi, Gholamreza Shahabadi, Alireza Moradi

Abstract: As CMOS technology scales into the deep nanometer regime, digital circuit reliability is increasingly threatened by the combined stochastic effects of Bias Temperature Instability (BTI) and Process Variation (PV). Traditional reliability analysis methods, which rely on computationally intensive simulations or extensive lookup tables, fail to scale efficiently for large designs, creating a critical… ▽ More As CMOS technology scales into the deep nanometer regime, digital circuit reliability is increasingly threatened by the combined stochastic effects of Bias Temperature Instability (BTI) and Process Variation (PV). Traditional reliability analysis methods, which rely on computationally intensive simulations or extensive lookup tables, fail to scale efficiently for large designs, creating a critical bottleneck in design space exploration. To address this, we propose SMART, a novel framework that integrates Machine Learning (ML) with Monte Carlo simulation to enable rapid, high-fidelity reliability analysis. SMART employs Random Forest regression to predict gate delay distributions directly, bypassing time-consuming atomic model parameter extractions. Crucially, the model utilizes Bayesian Optimization for automated hyperparameter tuning, ensuring maximum predictive robustness across diverse libraries. Experimental validation on ISCAS85 benchmark circuits demonstrates that SMART achieves a 94.54% reduction in analysis time compared to state-of-the-art methods, while maintaining a remarkable average accuracy error of just 1.63%. By shifting computational complexity to an offline training phase, the proposed framework offers a scalable, accurate solution for designing resilient, reliability-aware digital systems. △ Less

Submitted 7 July, 2026; v1 submitted 6 July, 2026; originally announced July 2026.

Comments: Submitted to Engineering Reports, Under Review

Federated Learning for Cross-Modality Medical Image Segmentation via Augmentation-Driven Generalization

Authors: Sachin Dudda Nagaraju, Ashkan Moradi, Bendik Skarre Abrahamsen, Mattijs Elschot

Abstract: Artificial intelligence has emerged as a transformative tool in medical image analysis, yet developing robust and generalizable segmentation models remains difficult due to fragmented, privacy-constrained imaging data siloed across institutions. While federated learning (FL) enables collaborative model training without centralizing data, cross-modality domain shifts pose a critical challenge, part… ▽ More Artificial intelligence has emerged as a transformative tool in medical image analysis, yet developing robust and generalizable segmentation models remains difficult due to fragmented, privacy-constrained imaging data siloed across institutions. While federated learning (FL) enables collaborative model training without centralizing data, cross-modality domain shifts pose a critical challenge, particularly when models trained on one modality fail to generalize to another. Many existing solutions require paired multimodal data per patient or rely on complex architectures, both of which are impractical in real clinical settings. In this work, we consider a realistic FL scenario where each client holds single-modality data (CT or MRI), and systematically investigate augmentation strategies for cross-modality generalization. Using abdominal organ segmentation and whole-heart segmentation as representative multi-class and binary segmentation benchmarks, we evaluate convolution-based spatial augmentation, frequency-domain manipulation, domain-specific normalization, and global intensity nonlinear (GIN) augmentation. Our results show that GIN consistently outperforms alternatives in both centralized and federated settings by simulating cross-modality appearance variations while preserving anatomical structure. For the pancreas, Dice score improved from 0.073 to 0.437, a 498% gain. Our federated approach achieves 93-98% of centralized training accuracy, demonstrating strong cross-modality generalization without compromising data privacy, pointing toward feasible federated AI deployment across diverse healthcare systems. △ Less

Submitted 24 February, 2026; originally announced February 2026.

Comments: Submitted to IEEE JBHI

Copula-Based Aggregation and Context-Aware Conformal Prediction for Reliable Renewable Energy Forecasting

Authors: Alireza Moradi, Mathieu Tanneau, Reza Zandehshahvar, Pascal Van Hentenryck

Abstract: The rapid growth of renewable energy penetration has intensified the need for reliable probabilistic forecasts to support grid operations at aggregated (fleet or system) levels. In practice, however, system operators often lack access to fleet-level probabilistic models and instead rely on site-level forecasts produced by heterogeneous third-party providers. Constructing coherent and calibrated fl… ▽ More The rapid growth of renewable energy penetration has intensified the need for reliable probabilistic forecasts to support grid operations at aggregated (fleet or system) levels. In practice, however, system operators often lack access to fleet-level probabilistic models and instead rely on site-level forecasts produced by heterogeneous third-party providers. Constructing coherent and calibrated fleet-level probabilistic forecasts from such inputs remains challenging due to complex cross-site dependencies and aggregation-induced miscalibration. This paper proposes a calibrated probabilistic aggregation framework that directly converts site-level probabilistic forecasts into reliable fleet-level forecasts in settings where system-level models cannot be trained or maintained. The framework integrates copula-based dependence modeling to capture cross-site correlations with Context-Aware Conformal Prediction (CACP) to correct miscalibration at the aggregated level. This combination enables dependence-aware aggregation while providing valid coverage and maintaining sharp prediction intervals. Experiments on large-scale solar generation datasets from MISO, ERCOT, and SPP demonstrate that the proposed Copula+CACP approach consistently achieves near-nominal coverage with significantly sharper intervals than uncalibrated aggregation baselines. △ Less

Submitted 31 January, 2026; originally announced February 2026.

Rural Connectivity Inequalities in Finland and Sweden: Evidence, Measures, and Policy Reflections

Authors: Sameera Bandaranayake, Amirreza Moradi, Tanja Suomalainen, Harri Saarnisaari, Pasi Karppinen, Payal Gupta, Jaap van de Beek

Abstract: Persistent rural-urban disparities in broadband connectivity remain a major policy challenge, even in digitally advanced countries. This paper examines how these inequalities manifest in northern Finland and Sweden, where sparse populations, long distances, and seasonal variations in demand create persistent gaps in service quality and reliability. Drawing on survey data (n = 148), in-depth interv… ▽ More Persistent rural-urban disparities in broadband connectivity remain a major policy challenge, even in digitally advanced countries. This paper examines how these inequalities manifest in northern Finland and Sweden, where sparse populations, long distances, and seasonal variations in demand create persistent gaps in service quality and reliability. Drawing on survey data (n = 148), in-depth interviews, and spatial analysis, the study explores the lived experience of connectivity in Arctic rural communities and introduces a novel Cellular Coverage Inequality (CCI) Index. The index combines measures of rurality and network performance to quantify spatial disparities that are masked by national coverage statistics. Results reveal that headline indicators overstate inclusiveness, while local users report chronic connectivity gaps affecting work, safety, and access to services. Building on these findings, the paper outlines policy reflections in six areas: shared infrastructure and roaming frameworks, spectrum flexibility for rural operators, performance-based Quality-of-Service monitoring, standardized and transparent reporting, temporal and seasonal capacity management, and digital-skills initiatives. Together, these recommendations highlight the need for multidimensional metrics and governance mechanisms that link technical performance, spatial equity, and user experience. The analysis contributes to ongoing debates on how broadband policy in sparsely populated regions can move beyond nominal coverage targets toward genuine inclusion and reliability. △ Less

Submitted 2 December, 2025; originally announced December 2025.

Enhanced Renewable Energy Forecasting using Context-Aware Conformal Prediction

Authors: Alireza Moradi, Mathieu Tanneau, Reza Zandehshahvar, Pascal Van Hentenryck

Abstract: Artificial intelligence (AI) is increasingly used to support renewable energy forecasting and grid operations. As renewable penetration grows, reliable probabilistic forecasting is becoming essential for managing uncertainty and supporting risk-aware operational decision-making. However, these forecasts often suffer from miscalibration due to temporal variability, changing weather conditions, and… ▽ More Artificial intelligence (AI) is increasingly used to support renewable energy forecasting and grid operations. As renewable penetration grows, reliable probabilistic forecasting is becoming essential for managing uncertainty and supporting risk-aware operational decision-making. However, these forecasts often suffer from miscalibration due to temporal variability, changing weather conditions, and heterogeneous operating regimes. In many real-world settings, renewable energy forecasts are provided by external sources, vendors, or independently trained systems, making retraining infeasible because of limited model access or computational constraints. This creates a need for efficient and model-agnostic methods that can improve forecast reliability after they are produced. This paper presents Context-Aware Conformal Prediction (CACP), a framework for calibrating renewable energy forecasts. The proposed method relies on a weighting mechanism during the calibration procedure which assigns higher weights to historical observations that are more similar to the target forecasting condition. This enables adaptive prediction intervals that reflect local uncertainty regimes without requiring access to, or retraining of, the underlying forecasting model. Experiments are performed on a large-scale dataset from National Renewable Energy Laboratory (NREL) day-ahead solar forecasting, covering multiple systems including MISO, ERCTO, and SPP. The results show that CACP improves the reliability-efficiency tradeoff at both site and system levels compared to NREL's base forecasting model and the other conformal prediction baselines. These results suggest that CACP can serve as a practical reliability-enhancement layer for trustworthy AI-enabled renewable energy forecasting and operational decision support. △ Less

Submitted 1 June, 2026; v1 submitted 17 October, 2025; originally announced October 2025.

FedGIN: Federated Learning with Dynamic Global Intensity Non-linear Augmentation for Organ Segmentation using Multi-modal Images

Authors: Sachin Dudda Nagaraju, Ashkan Moradi, Bendik Skarre Abrahamsen, Mattijs Elschot

Abstract: Medical image segmentation plays a crucial role in AI-assisted diagnostics, surgical planning, and treatment monitoring. Accurate and robust segmentation models are essential for enabling reliable, data-driven clinical decision making across diverse imaging modalities. Given the inherent variability in image characteristics across modalities, developing a unified model capable of generalizing effe… ▽ More Medical image segmentation plays a crucial role in AI-assisted diagnostics, surgical planning, and treatment monitoring. Accurate and robust segmentation models are essential for enabling reliable, data-driven clinical decision making across diverse imaging modalities. Given the inherent variability in image characteristics across modalities, developing a unified model capable of generalizing effectively to multiple modalities would be highly beneficial. This model could streamline clinical workflows and reduce the need for modality-specific training. However, real-world deployment faces major challenges, including data scarcity, domain shift between modalities (e.g., CT vs. MRI), and privacy restrictions that prevent data sharing. To address these issues, we propose FedGIN, a Federated Learning (FL) framework that enables multimodal organ segmentation without sharing raw patient data. Our method integrates a lightweight Global Intensity Non-linear (GIN) augmentation module that harmonizes modality-specific intensity distributions during local training. We evaluated FedGIN using two types of datasets: an imputed dataset and a complete dataset. In the limited dataset scenario, the model was initially trained using only MRI data, and CT data was added to assess its performance improvements. In the complete dataset scenario, both MRI and CT data were fully utilized for training on all clients. In the limited-data scenario, FedGIN achieved a 12 to 18% improvement in 3D Dice scores on MRI test cases compared to FL without GIN and consistently outperformed local baselines. In the complete dataset scenario, FedGIN demonstrated near-centralized performance, with a 30% Dice score improvement over the MRI-only baseline and a 10% improvement over the CT-only baseline, highlighting its strong cross-modality generalization under privacy constraints. △ Less

Submitted 7 August, 2025; originally announced August 2025.

Comments: Paper Accepted at MICCAI 2025 DeCaf Workshop Track

Journal ref: MICCAI 2025

Informed Forecasting: Leveraging Auxiliary Knowledge to Boost LLM Performance on Time Series Forecasting

Authors: Mohammadmahdi Ghasemloo, Alireza Moradi

Abstract: With the widespread adoption of Large Language Models (LLMs), there is a growing need to establish best practices for leveraging their capabilities beyond traditional natural language tasks. In this paper, a novel cross-domain knowledge transfer framework is proposed to enhance the performance of LLMs in time series forecasting -- a task of increasing relevance in fields such as energy systems, fi… ▽ More With the widespread adoption of Large Language Models (LLMs), there is a growing need to establish best practices for leveraging their capabilities beyond traditional natural language tasks. In this paper, a novel cross-domain knowledge transfer framework is proposed to enhance the performance of LLMs in time series forecasting -- a task of increasing relevance in fields such as energy systems, finance, and healthcare. The approach systematically infuses LLMs with structured temporal information to improve their forecasting accuracy. This study evaluates the proposed method on a real-world time series dataset and compares it to a naive baseline where the LLM receives no auxiliary information. Results show that knowledge-informed forecasting significantly outperforms the uninformed baseline in terms of predictive accuracy and generalization. These findings highlight the potential of knowledge transfer strategies to bridge the gap between LLMs and domain-specific forecasting tasks. △ Less

Submitted 25 September, 2025; v1 submitted 15 May, 2025; originally announced May 2025.

Federated Smoothing ADMM for Localization

Authors: Reza Mirzaeifard, Ashkan Moradi, Masahiro Yukawa, Stefan Werner

Abstract: This paper addresses the challenge of localization in federated settings, which are characterized by distributed data, non-convexity, and non-smoothness. To tackle the scalability and outlier issues inherent in such environments, we propose a robust algorithm that employs an $\ell_1$-norm formulation within a novel federated ADMM framework. This approach addresses the problem by integrating an ite… ▽ More This paper addresses the challenge of localization in federated settings, which are characterized by distributed data, non-convexity, and non-smoothness. To tackle the scalability and outlier issues inherent in such environments, we propose a robust algorithm that employs an $\ell_1$-norm formulation within a novel federated ADMM framework. This approach addresses the problem by integrating an iterative smooth approximation for the total variation consensus term and employing a Moreau envelope approximation for the convex function that appears in a subtracted form. This transformation ensures that the problem is smooth and weakly convex in each iteration, which results in enhanced computational efficiency and improved estimation accuracy. The proposed algorithm supports asynchronous updates and multiple client updates per iteration, which ensures its adaptability to real-world federated systems. To validate the reliability of the proposed algorithm, we show that the method converges to a stationary point, and numerical simulations highlight its superior performance in convergence speed and outlier resilience compared to existing state-of-the-art localization methods. △ Less

Submitted 12 March, 2025; originally announced March 2025.

How Low Can We Go? Minimizing Interaction Samples for Configurable Systems

Authors: Dominik Krupke, Ahmad Moradi, Michael Perk, Phillip Keldenich, Gabriel Gehrke, Sebastian Krieter, Thomas Thüm, Sándor P. Fekete

Abstract: Modern software systems are typically configurable, a fundamental prerequisite for wide applicability and reusability. This flexibility poses an extraordinary challenge for quality assurance, as the enormous number of possible configurations makes it impractical to test each of them separately. This is where t-wise interaction sampling can be used to systematically cover the configuration space an… ▽ More Modern software systems are typically configurable, a fundamental prerequisite for wide applicability and reusability. This flexibility poses an extraordinary challenge for quality assurance, as the enormous number of possible configurations makes it impractical to test each of them separately. This is where t-wise interaction sampling can be used to systematically cover the configuration space and detect unknown feature interactions. Over the last two decades, numerous algorithms for computing small interaction samples have been studied, providing improvements for a range of heuristic results; nevertheless, it has remained unclear how much these results can still be improved. We present a significant breakthrough: a fundamental framework, based on the mathematical principle of duality, for combining near-optimal solutions with provable lower bounds on the required sample size. This implies that we no longer need to work on heuristics with marginal or no improvement, but can certify the solution quality by establishing a limit on the remaining gap; in many cases, we can even prove optimality of achieved solutions. This theoretical contribution also provides extensive practical improvements: Our algorithm SampLNS was tested on 47 small and medium-sized configurable systems from the existing literature. SampLNS can reliably find samples of smaller size than previous methods in 85% of the cases; moreover, we can achieve and prove optimality of solutions for 63% of all instances. This makes it possible to avoid cumbersome efforts of minimizing samples by researchers as well as practitioners, and substantially save testing resources for most configurable systems. △ Less

Submitted 12 January, 2025; originally announced January 2025.

ACM Class: D.2; F.2.2

Embedding Cultural Diversity in Prototype-based Recommender Systems

Authors: Armin Moradi, Nicola Neophytou, Florian Carichon, Golnoosh Farnadi

Abstract: Popularity bias in recommender systems can increase cultural overrepresentation by favoring norms from dominant cultures and marginalizing underrepresented groups. This issue is critical for platforms offering cultural products, as they influence consumption patterns and human perceptions. In this work, we address popularity bias by identifying demographic biases within prototype-based matrix fact… ▽ More Popularity bias in recommender systems can increase cultural overrepresentation by favoring norms from dominant cultures and marginalizing underrepresented groups. This issue is critical for platforms offering cultural products, as they influence consumption patterns and human perceptions. In this work, we address popularity bias by identifying demographic biases within prototype-based matrix factorization methods. Using the country of origin as a proxy for cultural identity, we link this demographic attribute to popularity bias by refining the embedding space learning process. First, we propose filtering out irrelevant prototypes to improve representativity. Second, we introduce a regularization technique to enforce a uniform distribution of prototypes within the embedding space. Across four datasets, our results demonstrate a 27\% reduction in the average rank of long-tail items and a 2\% reduction in the average rank of items from underrepresented countries. Additionally, our model achieves a 2\% improvement in HitRatio@10 compared to the state-of-the-art, highlighting that fairness is enhanced without compromising recommendation quality. Moreover, the distribution of prototypes leads to more inclusive explanations by better aligning items with diverse prototypes. △ Less

Submitted 18 December, 2024; originally announced December 2024.

Large Language Models versus Classical Machine Learning: Performance in COVID-19 Mortality Prediction Using High-Dimensional Tabular Data

Authors: Mohammadreza Ghaffarzadeh-Esfahani, Mahdi Ghaffarzadeh-Esfahani, Arian Salahi-Niri, Hossein Toreyhi, Zahra Atf, Amirali Mohsenzadeh-Kermani, Mahshad Sarikhani, Zohreh Tajabadi, Fatemeh Shojaeian, Mohammad Hassan Bagheri, Aydin Feyzi, Mohammadamin Tarighatpayma, Narges Gazmeh, Fateme Heydari, Hossein Afshar, Amirreza Allahgholipour, Farid Alimardani, Ameneh Salehi, Naghmeh Asadimanesh, Mohammad Amin Khalafi, Hadis Shabanipour, Ali Moradi, Sajjad Hossein Zadeh, Omid Yazdani, Romina Esbati , et al. (17 additional authors not shown)

Abstract: This study compared the performance of classical feature-based machine learning models (CMLs) and large language models (LLMs) in predicting COVID-19 mortality using high-dimensional tabular data from 9,134 patients across four hospitals. Seven CML models, including XGBoost and random forest (RF), were evaluated alongside eight LLMs, such as GPT-4 and Mistral-7b, which performed zero-shot classifi… ▽ More This study compared the performance of classical feature-based machine learning models (CMLs) and large language models (LLMs) in predicting COVID-19 mortality using high-dimensional tabular data from 9,134 patients across four hospitals. Seven CML models, including XGBoost and random forest (RF), were evaluated alongside eight LLMs, such as GPT-4 and Mistral-7b, which performed zero-shot classification on text-converted structured data. Additionally, Mistral- 7b was fine-tuned using the QLoRA approach. XGBoost and RF demonstrated superior performance among CMLs, achieving F1 scores of 0.87 and 0.83 for internal and external validation, respectively. GPT-4 led the LLM category with an F1 score of 0.43, while fine-tuning Mistral-7b significantly improved its recall from 1% to 79%, yielding a stable F1 score of 0.74 during external validation. Although LLMs showed moderate performance in zero-shot classification, fine-tuning substantially enhanced their effectiveness, potentially bridging the gap with CML models. However, CMLs still outperformed LLMs in handling high-dimensional tabular data tasks. This study highlights the potential of both CMLs and fine-tuned LLMs in medical predictive modeling, while emphasizing the current superiority of CMLs for structured data analysis. △ Less

Submitted 26 September, 2025; v1 submitted 2 September, 2024; originally announced September 2024.

Comments: Code is available at: https://github.com/mohammad-gh009/Large-Language-Models-vs-Classical-Machine-learning and https://github.com/Sdamirsa/Tehran_COVID_Cohort. The datasets are available from the corresponding author on reasonable request (sdamirsa@ymail.com)

MSC Class: 92C50; 68T50 ACM Class: J.3

Journal ref: Scientific Reports 15, 42712 (2025)

Advancing Cultural Inclusivity: Optimizing Embedding Spaces for Balanced Music Recommendations

Authors: Armin Moradi, Nicola Neophytou, Golnoosh Farnadi

Abstract: Popularity bias in music recommendation systems -- where artists and tracks with the highest listen counts are recommended more often -- can also propagate biases along demographic and cultural axes. In this work, we identify these biases in recommendations for artists from underrepresented cultural groups in prototype-based matrix factorization methods. Unlike traditional matrix factorization met… ▽ More Popularity bias in music recommendation systems -- where artists and tracks with the highest listen counts are recommended more often -- can also propagate biases along demographic and cultural axes. In this work, we identify these biases in recommendations for artists from underrepresented cultural groups in prototype-based matrix factorization methods. Unlike traditional matrix factorization methods, prototype-based approaches are interpretable. This allows us to directly link the observed bias in recommendations for minority artists (the effect) to specific properties of the embedding space (the cause). We mitigate popularity bias in music recommendation through capturing both users' and songs' cultural nuances in the embedding space. To address these challenges while maintaining recommendation quality, we propose two novel enhancements to the embedding space: i) we propose an approach to filter-out the irrelevant prototypes used to represent each user and item to improve generalizability, and ii) we introduce regularization techniques to reinforce a more uniform distribution of prototypes within the embedding space. Our results demonstrate significant improvements in reducing popularity bias and enhancing demographic and cultural fairness in music recommendations while achieving competitive -- if not better -- overall performance. △ Less

Submitted 27 May, 2024; originally announced May 2024.

Partially-Precise Computing Paradigm for Efficient Hardware Implementation of Application-Specific Embedded Systems

Authors: Mohsen Faryabi, Amir Hossein Moradi

Abstract: Nowadays, the number of emerging embedded systems rapidly grows in many application domains, due to recent advances in artificial intelligence and internet of things. The main inherent specification of these application-specific systems is that they have not a general nature and are basically developed to only perform a particular task and therefore, deal only with a limited and predefined range o… ▽ More Nowadays, the number of emerging embedded systems rapidly grows in many application domains, due to recent advances in artificial intelligence and internet of things. The main inherent specification of these application-specific systems is that they have not a general nature and are basically developed to only perform a particular task and therefore, deal only with a limited and predefined range of custom input values. Despite this significant feature, these emerging applications are still conventionally implemented using general-purpose and precise digital computational blocks, which are essentially developed to provide the correct result for all possible input values. This highly degrades the physical properties of these applications while does not improve their functionality. To resolve this conflict, a novel computational paradigm named as partially-precise computing is introduced in this paper, based on an inspiration from the brain information reduction hypothesis as a tenet of neuroscience. The main specification of a Partially-Precise Computational (PPC) block is that it provides the precise result only for a desired, limited, and predefined set of input values. This relaxes its internal structure which results in improved physical properties with respect to a conventional precise block. The PPC blocks improve the implementation costs of the embedded applications, with a negligible or even without any output quality degradation with respect to the conventional implementation. The applicability and efficiency of the first instances of PPC adders and multipliers in a Gaussian denoising filter, an image blending and a face recognition neural network are demonstrated by means of a wide range of simulation and synthesis results. △ Less

Submitted 25 March, 2024; originally announced March 2024.

Comments: main article is 12 pages and supplementary notes is 6 pages

JustSTART: How to Find an RSA Authentication Bypass on Xilinx UltraScale(+) with Fuzzing

Authors: Maik Ender, Felix Hahn, Marc Fyrbiak, Amir Moradi, Christof Paar

Abstract: Fuzzing is a well-established technique in the software domain to uncover bugs and vulnerabilities. Yet, applications of fuzzing for security vulnerabilities in hardware systems are scarce, as principal reasons are requirements for design information access (HDL source code). Moreover, observation of internal hardware state during runtime is typically an ineffective information source, as its docu… ▽ More Fuzzing is a well-established technique in the software domain to uncover bugs and vulnerabilities. Yet, applications of fuzzing for security vulnerabilities in hardware systems are scarce, as principal reasons are requirements for design information access (HDL source code). Moreover, observation of internal hardware state during runtime is typically an ineffective information source, as its documentation is often not publicly available. In addition, such observation during runtime is also inefficient due to bandwidth-limited analysis interfaces (JTAG, and minimal introspection of internal modules). In this work, we investigate fuzzing for 7-Series and UltraScale(+) FPGA configuration engines, the control plane governing the (secure) bitstream configuration within the FPGA. Our goal is to examine the effectiveness of fuzzing to analyze and document the opaque inner workings of FPGA configuration engines, with a primary emphasis on identifying security vulnerabilities. Using only the publicly available chip and dispersed documentation, we first design and implement ConFuzz, an advanced FPGA configuration engine fuzzing and rapid prototyping framework. Based on our detailed understanding of the bitstream file format, we then systematically define 3 novel key fuzzing strategies for Xilinx configuration engines. Moreover, our strategies are executed through mutational structure-aware fuzzers and incorporate various novel custom-tailored, FPGA-specific optimizations. Our evaluation reveals previously undocumented behavior within the configuration engine, including critical findings such as system crashes leading to unresponsive states of the FPGA. In addition, our investigations not only lead to the rediscovery of the starbleed attack but also uncover JustSTART (CVE-2023-20570), capable of circumventing RSA authentication for Xilinx UltraScale(+). Note that we also discuss countermeasures. △ Less

Submitted 15 February, 2024; originally announced February 2024.

Tidying Up the Conversational Recommender Systems' Biases

Authors: Armin Moradi, Golnoosh Farnadi

Abstract: The growing popularity of language models has sparked interest in conversational recommender systems (CRS) within both industry and research circles. However, concerns regarding biases in these systems have emerged. While individual components of CRS have been subject to bias studies, a literature gap remains in understanding specific biases unique to CRS and how these biases may be amplified or r… ▽ More The growing popularity of language models has sparked interest in conversational recommender systems (CRS) within both industry and research circles. However, concerns regarding biases in these systems have emerged. While individual components of CRS have been subject to bias studies, a literature gap remains in understanding specific biases unique to CRS and how these biases may be amplified or reduced when integrated into complex CRS models. In this paper, we provide a concise review of biases in CRS by surveying recent literature. We examine the presence of biases throughout the system's pipeline and consider the challenges that arise from combining multiple models. Our study investigates biases in classic recommender systems and their relevance to CRS. Moreover, we address specific biases in CRS, considering variations with and without natural language understanding capabilities, along with biases related to dialogue systems and language models. Through our findings, we highlight the necessity of adopting a holistic perspective when dealing with biases in complex CRS models. △ Less

Submitted 5 September, 2023; originally announced September 2023.

Machine learning-based characterization of hydrochar from biomass: Implications for sustainable energy and material production

Authors: Alireza Shafizadeh, Hossein Shahbeik, Shahin Rafiee, Aysooda Moradi, Mohammadreza Shahbaz, Meysam Madadi, Cheng Li, Wanxi Peng, Meisam Tabatabaei, Mortaza Aghbashlo

Abstract: Hydrothermal carbonization (HTC) is a process that converts biomass into versatile hydrochar without the need for prior drying. The physicochemical properties of hydrochar are influenced by biomass properties and processing parameters, making it challenging to optimize for specific applications through trial-and-error experiments. To save time and money, machine learning can be used to develop a m… ▽ More Hydrothermal carbonization (HTC) is a process that converts biomass into versatile hydrochar without the need for prior drying. The physicochemical properties of hydrochar are influenced by biomass properties and processing parameters, making it challenging to optimize for specific applications through trial-and-error experiments. To save time and money, machine learning can be used to develop a model that characterizes hydrochar produced from different biomass sources under varying reaction processing parameters. Thus, this study aims to develop an inclusive model to characterize hydrochar using a database covering a range of biomass types and reaction processing parameters. The quality and quantity of hydrochar are predicted using two models (decision tree regression and support vector regression). The decision tree regression model outperforms the support vector regression model in terms of forecast accuracy (R2 > 0.88, RMSE < 6.848, and MAE < 4.718). Using an evolutionary algorithm, optimum inputs are identified based on cost functions provided by the selected model to optimize hydrochar for energy production, soil amendment, and pollutant adsorption, resulting in hydrochar yields of 84.31%, 84.91%, and 80.40%, respectively. The feature importance analysis reveals that biomass ash/carbon content and operating temperature are the primary factors affecting hydrochar production in the HTC process. △ Less

Submitted 24 May, 2023; originally announced May 2023.

Journal ref: Fuel 347, 1 September 2023, 128467

Language Model Crossover: Variation through Few-Shot Prompting

Authors: Elliot Meyerson, Mark J. Nelson, Herbie Bradley, Adam Gaier, Arash Moradi, Amy K. Hoover, Joel Lehman

Abstract: This paper pursues the insight that language models naturally enable an intelligent variation operator similar in spirit to evolutionary crossover. In particular, language models of sufficient scale demonstrate in-context learning, i.e. they can learn from associations between a small number of input patterns to generate outputs incorporating such associations (also called few-shot prompting). Thi… ▽ More This paper pursues the insight that language models naturally enable an intelligent variation operator similar in spirit to evolutionary crossover. In particular, language models of sufficient scale demonstrate in-context learning, i.e. they can learn from associations between a small number of input patterns to generate outputs incorporating such associations (also called few-shot prompting). This ability can be leveraged to form a simple but powerful variation operator, i.e. to prompt a language model with a few text-based genotypes (such as code, plain-text sentences, or equations), and to parse its corresponding output as those genotypes' offspring. The promise of such language model crossover (which is simple to implement and can leverage many different open-source language models) is that it enables a simple mechanism to evolve semantically-rich text representations (with few domain-specific tweaks), and naturally benefits from current progress in language models. Experiments in this paper highlight the versatility of language-model crossover, through evolving binary bit-strings, sentences, equations, text-to-image prompts, and Python code. The conclusion is that language model crossover is a promising method for evolving genomes representable as text. △ Less

Submitted 13 May, 2024; v1 submitted 23 February, 2023; originally announced February 2023.

MCTS with Refinement for Proposals Selection Games in Scene Understanding

Authors: Sinisa Stekovic, Mahdi Rad, Alireza Moradi, Friedrich Fraundorfer, Vincent Lepetit

Abstract: We propose a novel method applicable in many scene understanding problems that adapts the Monte Carlo Tree Search (MCTS) algorithm, originally designed to learn to play games of high-state complexity. From a generated pool of proposals, our method jointly selects and optimizes proposals that minimize the objective term. In our first application for floor plan reconstruction from point clouds, our… ▽ More We propose a novel method applicable in many scene understanding problems that adapts the Monte Carlo Tree Search (MCTS) algorithm, originally designed to learn to play games of high-state complexity. From a generated pool of proposals, our method jointly selects and optimizes proposals that minimize the objective term. In our first application for floor plan reconstruction from point clouds, our method selects and refines the room proposals, modelled as 2D polygons, by optimizing on an objective function combining the fitness as predicted by a deep network and regularizing terms on the room shapes. We also introduce a novel differentiable method for rendering the polygonal shapes of these proposals. Our evaluations on the recent and challenging Structured3D and Floor-SP datasets show significant improvements over the state-of-the-art, without imposing hard constraints nor assumptions on the floor plan configurations. In our second application, we extend our approach to reconstruct general 3D room layouts from a color image and obtain accurate room layouts. We also show that our differentiable renderer can easily be extended for rendering 3D planar polygons and polygon embeddings. Our method shows high performance on the Matterport3D-Layout dataset, without introducing hard constraints on room layout configurations. △ Less

Submitted 7 July, 2022; originally announced July 2022.

Comments: Submitted to: TPAMI Special Section on the Best Papers of ICCV2021 GitHub Repository: https://github.com/vevenom/MonteScene. arXiv admin note: substantial text overlap with arXiv:2103.11161

The Klarna Product Page Dataset: Web Element Nomination with Graph Neural Networks and Large Language Models

Authors: Alexandra Hotti, Riccardo Sven Risuleo, Stefan Magureanu, Aref Moradi, Jens Lagergren

Abstract: Web automation holds the potential to revolutionize how users interact with the digital world, offering unparalleled assistance and simplifying tasks via sophisticated computational methods. Central to this evolution is the web element nomination task, which entails identifying unique elements on webpages. Unfortunately, the development of algorithmic designs for web automation is hampered by the… ▽ More Web automation holds the potential to revolutionize how users interact with the digital world, offering unparalleled assistance and simplifying tasks via sophisticated computational methods. Central to this evolution is the web element nomination task, which entails identifying unique elements on webpages. Unfortunately, the development of algorithmic designs for web automation is hampered by the scarcity of comprehensive and realistic datasets that reflect the complexity faced by real-world applications on the Web. To address this, we introduce the Klarna Product Page Dataset, a comprehensive and diverse collection of webpages that surpasses existing datasets in richness and variety. The dataset features 51,701 manually labeled product pages from 8,175 e-commerce websites across eight geographic regions, accompanied by a dataset of rendered page screenshots. To initiate research on the Klarna Product Page Dataset, we empirically benchmark a range of Graph Neural Networks (GNNs) on the web element nomination task. We make three important contributions. First, we found that a simple Convolutional GNN (GCN) outperforms complex state-of-the-art nomination methods. Second, we introduce a training refinement procedure that involves identifying a small number of relevant elements from each page using the aforementioned GCN. These elements are then passed to a large language model for the final nomination. This procedure significantly improves the nomination accuracy by 16.8 percentage points on our challenging dataset, without any need for fine-tuning. Finally, in response to another prevalent challenge in this field - the abundance of training methodologies suitable for element nomination - we introduce the Challenge Nomination Training Procedure, a novel training approach that further boosts nomination accuracy. △ Less

Submitted 23 February, 2024; v1 submitted 3 November, 2021; originally announced November 2021.

Comments: 12 pages, 8 figures, 3 tables, under review

MSC Class: 68T07

Journal ref: Transactions on Machine Learning Research, 2024

The Unpatchable Silicon: A Full Break of the Bitstream Encryption of Xilinx 7-Series FPGAs

Authors: Maik Ender, Amir Moradi, Christof Paar

Abstract: The security of FPGAs is a crucial topic, as any vulnerability within the hardware can have severe consequences, if they are used in a secure design. Since FPGA designs are encoded in a bitstream, securing the bitstream is of the utmost importance. Adversaries have many motivations to recover and manipulate the bitstream, including design cloning, IP theft, manipulation of the design, or design su… ▽ More The security of FPGAs is a crucial topic, as any vulnerability within the hardware can have severe consequences, if they are used in a secure design. Since FPGA designs are encoded in a bitstream, securing the bitstream is of the utmost importance. Adversaries have many motivations to recover and manipulate the bitstream, including design cloning, IP theft, manipulation of the design, or design subversions e.g., through hardware Trojans. Given that FPGAs are often part of cyber-physical systems e.g., in aviation, medical, or industrial devices, this can even lead to physical harm. Consequently, vendors have introduced bitstream encryption, offering authenticity and confidentiality. Even though attacks against bitstream encryption have been proposed in the past, e.g., side-channel analysis and probing, these attacks require sophisticated equipment and considerable technical expertise. In this paper, we introduce novel low-cost attacks against the Xilinx 7-Series (and Virtex-6) bitstream encryption, resulting in the total loss of authenticity and confidentiality. We exploit a design flaw which piecewise leaks the decrypted bitstream. In the attack, the FPGA is used as a decryption oracle, while only access to a configuration interface is needed. The attack does not require any sophisticated tools and, depending on the target system, can potentially be launched remotely. In addition to the attacks, we discuss several countermeasures. △ Less

Submitted 28 May, 2021; originally announced May 2021.

Real-World Snapshots vs. Theory: Questioning the t-Probing Security Model

Authors: Thilo Krachenfels, Fatemeh Ganji, Amir Moradi, Shahin Tajik, Jean-Pierre Seifert

Abstract: Due to its sound theoretical basis and practical efficiency, masking has become the most prominent countermeasure to protect cryptographic implementations against physical side-channel attacks (SCAs). The core idea of masking is to randomly split every sensitive intermediate variable during computation into at least t+1 shares, where t denotes the maximum number of shares that are allowed to be ob… ▽ More Due to its sound theoretical basis and practical efficiency, masking has become the most prominent countermeasure to protect cryptographic implementations against physical side-channel attacks (SCAs). The core idea of masking is to randomly split every sensitive intermediate variable during computation into at least t+1 shares, where t denotes the maximum number of shares that are allowed to be observed by an adversary without learning any sensitive information. In other words, it is assumed that the adversary is bounded either by the possessed number of probes (e.g., microprobe needles) or by the order of statistical analyses while conducting higher-order SCA attacks (e.g., differential power analysis). Such bounded models are employed to prove the SCA security of the corresponding implementations. Consequently, it is believed that given a sufficiently large number of shares, the vast majority of known SCA attacks are mitigated. In this work, we present a novel laser-assisted SCA technique, called Laser Logic State Imaging (LLSI), which offers an unlimited number of contactless probes, and therefore, violates the probing security model assumption. This technique enables us to take snapshots of hardware implementations, i.e., extract the logical state of all registers at any arbitrary clock cycle with a single measurement. To validate this, we mount our attack on masked AES hardware implementations and practically demonstrate the extraction of the full-length key in two different scenarios. First, we assume that the location of the registers (key and/or state) is known, and hence, their content can be directly read by a single snapshot. Second, we consider an implementation with unknown register locations, where we make use of multiple snapshots and a SAT solver to reveal the secrets. △ Less

Submitted 9 September, 2020; originally announced September 2020.

Comments: This is the authors' version of the article accepted for publication at IEEE Symposium on Security and Privacy 2021

Interdiction in Practice -- Hardware Trojan Against a High-Security USB Flash Drive

Authors: Pawel Swierczynski, Marc Fyrbiak, Philipp Koppe, Amir Moradi, Christof Paar

Abstract: As part of the revelations about the NSA activities, the notion of interdiction has become known to the public: the interception of deliveries to manipulate hardware in a way that backdoors are introduced. Manipulations can occur on the firmware or at hardware level. With respect to hardware, FPGAs are particular interesting targets as they can be altered by manipulating the corresponding bitstrea… ▽ More As part of the revelations about the NSA activities, the notion of interdiction has become known to the public: the interception of deliveries to manipulate hardware in a way that backdoors are introduced. Manipulations can occur on the firmware or at hardware level. With respect to hardware, FPGAs are particular interesting targets as they can be altered by manipulating the corresponding bitstream which configures the device. In this paper, we demonstrate the first successful real-world FPGA hardware Trojan insertion into a commercial product. On the target device, a FIPS-140-2 level 2 certified USB flash drive from Kingston, the user data are encrypted using AES-256 in XTS mode, and the encryption/decryption is processed by an off-the-shelf SRAM-based FPGA. Our investigation required two reverse-engineering steps, related to the proprietary FPGA bitstream and to the firmware of the underlying ARM CPU. In our Trojan insertion scenario, the targeted USB flash drive is intercepted before being delivered to the victim. The physical Trojan insertion requires the manipulation of the SPI flash memory content, which contains the FPGA bitstream as well as the ARM CPU code. The FPGA bitstream manipulation alters the exploited AES-256 algorithm in a way that it turns into a linear function which can be broken with 32 known plaintext-ciphertext pairs. After the manipulated USB flash drive has been used by the victim, the attacker is able to obtain all user data from the ciphertexts. Our work indeed highlights the security risks and especially the practical relevance of bitstream modification attacks that became realistic due to FPGA bitstream manipulations. △ Less

Submitted 1 October, 2019; originally announced October 2019.

Journal ref: J Cryptogr Eng (2017) 7: 199

Side-Channel Hardware Trojan for Provably-Secure SCA-Protected Implementations

Authors: Samaneh Ghandali, Thorben Moos, Amir Moradi, Christof Paar

Abstract: Hardware Trojans have drawn the attention of academia, industry and government agencies. Effective detection mechanisms and countermeasures against such malicious designs can only be developed when there is a deep understanding of how hardware Trojans can be built in practice, in particular Trojans specifically designed to avoid detection. In this work, we present a mechanism to introduce an extre… ▽ More Hardware Trojans have drawn the attention of academia, industry and government agencies. Effective detection mechanisms and countermeasures against such malicious designs can only be developed when there is a deep understanding of how hardware Trojans can be built in practice, in particular Trojans specifically designed to avoid detection. In this work, we present a mechanism to introduce an extremely stealthy hardware Trojan into cryptographic primitives equipped with provably-secure first-order side-channel countermeasures. Once the Trojan is triggered, the malicious design exhibits exploitable side-channel leakage, leading to successful key recovery attacks. Generally, such a Trojan requires neither addition nor removal of any logic which makes it extremely hard to detect. On ASICs, it can be inserted by subtle manipulations at the sub-transistor level and on FPGAs by changing the routing of particular signals, leading to \textbf{zero} logic overhead. The underlying concept is based on modifying a securely-masked hardware implementation in such a way that running the device at a particular clock frequency violates one of its essential properties, leading to exploitable leakage. We apply our technique to a Threshold Implementation of the PRESENT block cipher realized in two different CMOS technologies, and show that triggering the Trojan makes the ASIC prototypes vulnerable. △ Less

Submitted 22 September, 2019; originally announced October 2019.

Easily implementable time series forecasting techniques for resource provisioning in cloud computing

Authors: Michel Fliess, Cédric Join, Maria Bekcheva, Alireza Moradi, Hugues Mounier

Abstract: Workload predictions in cloud computing is obviously an important topic. Most of the existing publications employ various time series techniques, that might be difficult to implement. We suggest here another route, which has already been successfully used in financial engineering and photovoltaic energy. No mathematical modeling and machine learning procedures are needed. Our computer simulations… ▽ More Workload predictions in cloud computing is obviously an important topic. Most of the existing publications employ various time series techniques, that might be difficult to implement. We suggest here another route, which has already been successfully used in financial engineering and photovoltaic energy. No mathematical modeling and machine learning procedures are needed. Our computer simulations via realistic data, which are quite convincing, show that a setting mixing algebraic estimation techniques and the daily seasonality behaves much better. An application to the computing resource allocation, via virtual machines, is sketched out. △ Less

Submitted 11 March, 2019; v1 submitted 20 February, 2019; originally announced March 2019.

Comments: 6th Conference on Control, Decision and Information Technologies (CoDIT 2019), April 2019, Paris

Improving resource elasticity in cloud computing thanks to model-free control

Authors: Maria Bekcheva, Michel Fliess, Cédric Join, Alireza Moradi, Hugues Mounier

Abstract: In cloud computing management, the dynamic adaptation of computing resource allocations under time-varying workload is an active domain of investigation. Several control strategies were already proposed. Here the model-free control setting and the corresponding "intelligent" controllers, which are most successful in many concrete engineering situations, are employed for the "horizontal elasticity.… ▽ More In cloud computing management, the dynamic adaptation of computing resource allocations under time-varying workload is an active domain of investigation. Several control strategies were already proposed. Here the model-free control setting and the corresponding "intelligent" controllers, which are most successful in many concrete engineering situations, are employed for the "horizontal elasticity." When compared to the commercial "Auto-Scaling" algorithms, our easily implementable approach, behaves better even with sharp workload fluctuations. This is confirmed by experiments on Amazon Web Services (AWS). △ Less

Submitted 8 October, 2018; originally announced October 2018.

Comments: in French

Journal ref: ISTE OpenScience Automatique/Control, vol. 2, 2018