Gaël Richard
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Gaël Richard has a total of 90 co-authors. Here are the top co-authors by number of shared publications:
- Roland Badeau
Laboratoire Traitement et Communication de l'Information, Télécom Paris, Institut Polytechnique de Paris , Palaiseau 91120
11 shared publications - Slim Essid
Institut Telecom Paris, CNRS-LTCI, 91120 Palaiseau, France
4 shared publications - Umut Şimşekli
Telecom ParisTech, Paris, France
4 shared publications - Alexey Ozerov
Technicolor
3 shared publications - Sanjeel Parekh
Laboratoire de Traitement et Communication de lInformation (LTCI), Tlcom Paris, Institut Polytechnique de Paris, Palaiseau, France
3 shared publications - Ali Taylan Cemgil
Department of Computer Engineering, Boğaziçi University, İstanbul, Turkey
2 shared publications
Latest Publications
- 19 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 2215-2219
AudioCAN: Enhanced Few-Shot Audio Classification via Energy-Guided Temporal Cross Attention
Despite the growing research interest and practical potential of few-shot audio classification, its efficacy remains limited by the temporal sparsity of target sound events and the presence of non-target acoustic interference within audio samples. In this work, we leverage the cross attention mechanism to address these challenges. While the Cross Attention Network (CAN) has demonstrated superior performance in few-shot image classification by highlighting semantically relevant regions between support and query features, we propose AudioCAN by extending CAN to audio domain through two-fold, audio-specific modifications. In particular, we first reformulate the original 2D spatial cross attention to 1D temporal version to prioritize the time frames containing discriminative acoustic information, thereby mitigating interference from non-target sounds. Secondly, we introduce an energy-guided masking strategy to synthesize a pseudo-query set from the support set via data augmentation, which then serves as a guidance for optimizing the 1D temporal cross attention during training. Experimental results on several few-shot audio classification benchmarks demonstrate that AudioCAN achieves state-of-the-art performance in 5-way-1-shot settings while remaining highly competitive in 5-way-5-shot configurations.
- 19 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 2215-2219
AudioCAN: Enhanced Few-Shot Audio Classification via Energy-Guided Temporal Cross Attention
Despite the growing research interest and practical potential of few-shot audio classification, its efficacy remains limited by the temporal sparsity of target sound events and the presence of non-target acoustic interference within audio samples. In this work, we leverage the cross attention mechanism to address these challenges. While the Cross Attention Network (CAN) has demonstrated superior performance in few-shot image classification by highlighting semantically relevant regions between support and query features, we propose AudioCAN by extending CAN to audio domain through two-fold, audio-specific modifications. In particular, we first reformulate the original 2D spatial cross attention to 1D temporal version to prioritize the time frames containing discriminative acoustic information, thereby mitigating interference from non-target sounds. Secondly, we introduce an energy-guided masking strategy to synthesize a pseudo-query set from the support set via data augmentation, which then serves as a guidance for optimizing the 1D temporal cross attention during training. Experimental results on several few-shot audio classification benchmarks demonstrate that AudioCAN achieves state-of-the-art performance in 5-way-1-shot settings while remaining highly competitive in 5-way-5-shot configurations.
- Open Access
- 31 March 2026
- Published by Springer Nature in Journal on Audio, Speech, and Music Processing
- Vol. 2026 (1)
The Hi-Audio online platform for recording and distributing multi-track music datasets
This paper introduces the Hi-Audio online platform, an open-source tool designed to support musicians and researchers in the field of Music Information Retrieval (MIR). The platform enables the recording, uploading, and sharing of multitrack musical compositions, aiming to build an open-access audio database to advance research in music technology. Uploaded audio files are automatically analyzed upon synchronization with the server, leveraging signal processing techniques and machine learning models to generate rich metadata. The platform facilitates remote and asynchronous collaboration via a web-based interface accessible at hiaudio.fr. Furthermore, a novel built-in method for accurate and robust round-trip latency estimation in the browser is proposed and integrated into the platform, demonstrating its applicability in real-world distributed recording scenarios. Finally, an initial user evaluation with musicians was conducted to assess usability and practical relevance under realistic usage conditions. The evaluation combined task-based performance analysis with standardized usability and workload measures. The results indicate high task completion rates for core recording functions and show that the platform can be used effectively by musicians with minimal prior training.
- 31 March 2026
- Published by Springer Nature in Journal on Audio, Speech, and Music Processing
- Vol. 2026 (1)
Open AccessThe Hi-Audio online platform for recording and distributing multi-track music datasets
This paper introduces the Hi-Audio online platform, an open-source tool designed to support musicians and researchers in the field of Music Information Retrieval (MIR). The platform enables the recording, uploading, and sharing of multitrack musical compositions, aiming to build an open-access audio database to advance research in music technology. Uploaded audio files are automatically analyzed upon synchronization with the server, leveraging signal processing techniques and machine learning models to generate rich metadata. The platform facilitates remote and asynchronous collaboration via a web-based interface accessible at hiaudio.fr. Furthermore, a novel built-in method for accurate and robust round-trip latency estimation in the browser is proposed and integrated into the platform, demonstrating its applicability in real-world distributed recording scenarios. Finally, an initial user evaluation with musicians was conducted to assess usability and practical relevance under realistic usage conditions. The evaluation combined task-based performance analysis with standardized usability and workload measures. The results indicate high task completion rates for core recording functions and show that the platform can be used effectively by musicians with minimal prior training.
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 1552-1563
U-DREAM: Unsupervised Dereverberation Guided by a Reverberation Model
This paper explores the outcome of training state-of-the-art dereverberation models with supervision settings ranging from weakly-supervised to virtually unsupervised, relying solely on reverberant signals and an acoustic model for training. Most of the existing deep learning approaches typically require paired dry and reverberant data, which are difficult to obtain in practice. We develop instead a sequential learning strategy motivated by a maximum-likelihood formulation of the dereverberation problem, wherein acoustic parameters and dry signals are estimated from reverberant inputs using deep neural networks, guided by a reverberation matching loss. Our most data-efficient variant requires only 100 reverberation-parameter-labeled samples to outperform an unsupervised baseline, demonstrating the effectiveness and practicality of the proposed method in low-resource scenarios.
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 1552-1563
U-DREAM: Unsupervised Dereverberation Guided by a Reverberation Model
This paper explores the outcome of training state-of-the-art dereverberation models with supervision settings ranging from weakly-supervised to virtually unsupervised, relying solely on reverberant signals and an acoustic model for training. Most of the existing deep learning approaches typically require paired dry and reverberant data, which are difficult to obtain in practice. We develop instead a sequential learning strategy motivated by a maximum-likelihood formulation of the dereverberation problem, wherein acoustic parameters and dry signals are estimated from reverberant inputs using deep neural networks, guided by a reverberation matching loss. Our most data-efficient variant requires only 100 reverberation-parameter-labeled samples to outperform an unsupervised baseline, demonstrating the effectiveness and practicality of the proposed method in low-resource scenarios.
- 9 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 1388-1400
Of All StrIPEs: Investigating Structure-Informed Positional Encoding for Efficient Music Generation
While music remains a challenging domain for generative models like Transformers, a two-pronged approach has recently proved successful: inserting musically-relevant structural information into the positional encoding (PE) module and using kernel approximation techniques based on Random Fourier Features (RFF) to lower the computational cost from quadratic to linear. Yet, it is not clear how such RFF-based efficient PEs compare with those based on rotation matrices, such as Rotary Positional Encoding (RoPE). In this paper, we present a unified framework based on kernel methods to analyze both families of efficient PEs. We use this framework to develop a novel PE method called RoPEPool, capable of extracting causal relationships from temporal sequences. Using RFF-based PEs and rotation-based PEs, we demonstrate how seemingly disparate PEs can be jointly studied by considering the interactions they induce between two descriptive levels of the data: the input, capturing quickly-varying components, and the prior, capturing slowly-varying components. For empirical validation, we use a symbolic music generation task, namely, melody harmonization. We show that RoPEPool, combined with highly-informative structural priors, outperforms all methods.
- 9 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 1388-1400
Of All StrIPEs: Investigating Structure-Informed Positional Encoding for Efficient Music Generation
While music remains a challenging domain for generative models like Transformers, a two-pronged approach has recently proved successful: inserting musically-relevant structural information into the positional encoding (PE) module and using kernel approximation techniques based on Random Fourier Features (RFF) to lower the computational cost from quadratic to linear. Yet, it is not clear how such RFF-based efficient PEs compare with those based on rotation matrices, such as Rotary Positional Encoding (RoPE). In this paper, we present a unified framework based on kernel methods to analyze both families of efficient PEs. We use this framework to develop a novel PE method called RoPEPool, capable of extracting causal relationships from temporal sequences. Using RFF-based PEs and rotation-based PEs, we demonstrate how seemingly disparate PEs can be jointly studied by considering the interactions they induce between two descriptive levels of the data: the input, capturing quickly-varying components, and the prior, capturing slowly-varying components. For empirical validation, we use a symbolic music generation task, namely, melody harmonization. We show that RoPEPool, combined with highly-informative structural priors, outperforms all methods.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 1-12
Melody-Lyrics Matching with Contrastive Alignment Loss
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 1-12
Melody-Lyrics Matching with Contrastive Alignment Loss
- 5 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 84-95
The Inverse Drum Machine: Source Separation Through Joint Transcription and Analysis-by-Synthesis
We present the Inverse Drum Machine, a novel approach to Drum Source Separation that leverages an analysis-by-synthesis framework combined with deep learning. Unlike recent supervised methods that require isolated stem recordings for training, our approach is trained on drum mixtures with only transcription annotations. IDM integrates Automatic Drum Transcription and One-shot Drum Sample Synthesis, jointly optimizing these tasks in an end-to-end manner. By convolving synthesized one-shot samples with estimated onsets, akin to a drum machine, we reconstruct the individual drum stems and train a Deep Neural Network on the reconstruction of the mixture. Experiments on the StemGMD dataset demonstrate that IDM achieves separation quality comparable to state-of-the-art supervised methods that require isolated stems data.
- 5 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 34, 84-95
The Inverse Drum Machine: Source Separation Through Joint Transcription and Analysis-by-Synthesis
We present the Inverse Drum Machine, a novel approach to Drum Source Separation that leverages an analysis-by-synthesis framework combined with deep learning. Unlike recent supervised methods that require isolated stem recordings for training, our approach is trained on drum mixtures with only transcription annotations. IDM integrates Automatic Drum Transcription and One-shot Drum Sample Synthesis, jointly optimizing these tasks in an end-to-end manner. By convolving synthesized one-shot samples with estimated onsets, akin to a drum machine, we reconstruct the individual drum stems and train a Deep Neural Network on the reconstruction of the mixture. Experiments on the StemGMD dataset demonstrate that IDM achieves separation quality comparable to state-of-the-art supervised methods that require isolated stems data.
- Open Access
Melody-Lyrics Matching with Contrastive Alignment Loss
The connection between music and lyrics is far beyond semantic bonds. Conceptual pairs in the two modalities such as rhythm and rhyme, note duration and syllabic stress, and structure correspondence, raise a compelling yet seldom-explored direction in the field of music information retrieval. In this paper, we present melody-lyrics matching (MLM), a new task which retrieves potential lyrics for a given symbolic melody from text sources. Rather than generating lyrics from scratch, MLM essentially exploits the relationships between melody and lyrics. We propose a self-supervised representation learning framework with contrastive alignment loss for melody and lyrics. This has the potential to leverage the abundance of existing songs with paired melody and lyrics. No alignment annotations are required. Additionally, we introduce sylphone, a novel representation for lyrics at syllable-level activated by phoneme identity and vowel stress. We demonstrate that our method can match melody with coherent and singable lyrics with empirical results and intuitive examples. We open source code and provide matching examples on the companion webpage: https://github.com/changhongw/mlm.
Open AccessMelody-Lyrics Matching with Contrastive Alignment Loss
The connection between music and lyrics is far beyond semantic bonds. Conceptual pairs in the two modalities such as rhythm and rhyme, note duration and syllabic stress, and structure correspondence, raise a compelling yet seldom-explored direction in the field of music information retrieval. In this paper, we present melody-lyrics matching (MLM), a new task which retrieves potential lyrics for a given symbolic melody from text sources. Rather than generating lyrics from scratch, MLM essentially exploits the relationships between melody and lyrics. We propose a self-supervised representation learning framework with contrastive alignment loss for melody and lyrics. This has the potential to leverage the abundance of existing songs with paired melody and lyrics. No alignment annotations are required. Additionally, we introduce sylphone, a novel representation for lyrics at syllable-level activated by phoneme identity and vowel stress. We demonstrate that our method can match melody with coherent and singable lyrics with empirical results and intuitive examples. We open source code and provide matching examples on the companion webpage: https://github.com/changhongw/mlm.
- Open Access
A Maximum Length Sequence-Based Method for Robust Round-Trip Latency Estimation in online Digital Audio Workstations
Accurate estimation of latency when working with digital audio equipment is critical for the precise operation of certain applications. This is particularly true for Digital Audio Workstations (DAWs) and other tools used in the creation and editing of audio, especially music. These systems require exact synchronization or alignment of tracks, which is essential for the mixing process. Latency is an inherent phenomenon in digital audio capture, and although it may sometimes be minimal, it is always variable depending on the device, operating system, and audio configuration. The undesired effect introduced by latency-specifically referred to in this context as round-trip latency-manifests as a delay between the audio input and the corresponding output. The most effective way to address this issue is through prior measurement to enable proper compensation. Various methods exist for performing this measurement, generally based on the playback and recording of acoustic signals. This article presents an existing method applied in a novel way within the domain of audio and web browsers, based on the use of a Maximum Length Sequence (MLS) signal. This signal is commonly used in room impulse response characterization. To validate its effectiveness and identify the limitations of the proposed approach, multiple tests and experiments were conducted on different devices. Results were compared across various browsers and operating systems, and the proposed solution was benchmarked against the methods employed by existing online DAWs. The implementation of the proposed method is available as part of the Hi-Audio online platform-an open-source, browser-based DAW-providing a practical demonstration of its applicability and integration in real-world web audio environments.
Open AccessA Maximum Length Sequence-Based Method for Robust Round-Trip Latency Estimation in online Digital Audio Workstations
Accurate estimation of latency when working with digital audio equipment is critical for the precise operation of certain applications. This is particularly true for Digital Audio Workstations (DAWs) and other tools used in the creation and editing of audio, especially music. These systems require exact synchronization or alignment of tracks, which is essential for the mixing process. Latency is an inherent phenomenon in digital audio capture, and although it may sometimes be minimal, it is always variable depending on the device, operating system, and audio configuration. The undesired effect introduced by latency-specifically referred to in this context as round-trip latency-manifests as a delay between the audio input and the corresponding output. The most effective way to address this issue is through prior measurement to enable proper compensation. Various methods exist for performing this measurement, generally based on the playback and recording of acoustic signals. This article presents an existing method applied in a novel way within the domain of audio and web browsers, based on the use of a Maximum Length Sequence (MLS) signal. This signal is commonly used in room impulse response characterization. To validate its effectiveness and identify the limitations of the proposed approach, multiple tests and experiments were conducted on different devices. Results were compared across various browsers and operating systems, and the proposed solution was benchmarked against the methods employed by existing online DAWs. The implementation of the proposed method is available as part of the Hi-Audio online platform-an open-source, browser-based DAW-providing a practical demonstration of its applicability and integration in real-world web audio environments.
- Top Cited Papers
- 1 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Magazine
- Vol. 41 (6) , 51-59
Model-Based Deep Learning for Music Information Research: Leveraging diverse knowledge sources to enhance explainability, controllability, and resource efficiency
In this article, we investigate the notion of model-based deep learning in the realm of music information research (MIR). Loosely speaking, we refer to the term model-based deep learning for approaches that combine traditional knowledge-based methods with data-driven techniques, especially those based on deep learning, within a differentiable computing framework. In music, prior knowledge for instance related to sound production, music perception or music composition theory can be incorporated into the design of neural networks and associated loss functions. We outline three specific scenarios to illustrate the application of model-based deep learning in MIR, demonstrating the implementation of such concepts and their potential.
- 1 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Magazine
- Vol. 41 (6) , 51-59
Top Cited PapersModel-Based Deep Learning for Music Information Research: Leveraging diverse knowledge sources to enhance explainability, controllability, and resource efficiency
In this article, we investigate the notion of model-based deep learning in the realm of music information research (MIR). Loosely speaking, we refer to the term model-based deep learning for approaches that combine traditional knowledge-based methods with data-driven techniques, especially those based on deep learning, within a differentiable computing framework. In music, prior knowledge for instance related to sound production, music perception or music composition theory can be incorporated into the design of neural networks and associated loss functions. We outline three specific scenarios to illustrate the application of model-based deep learning in MIR, demonstrating the implementation of such concepts and their potential.
- Open Access
Annealed Multiple Choice Learning: Overcoming limitations of Winner-takes-all with annealing
We introduce Annealed Multiple Choice Learning (aMCL) which combines simulated annealing with MCL. MCL is a learning framework handling ambiguous tasks by predicting a small set of plausible hypotheses. These hypotheses are trained using the Winner-takes-all (WTA) scheme, which promotes the diversity of the predictions. However, this scheme may converge toward an arbitrarily suboptimal local minimum, due to the greedy nature of WTA. We overcome this limitation using annealing, which enhances the exploration of the hypothesis space during training. We leverage insights from statistical physics and information theory to provide a detailed description of the model training trajectory. Additionally, we validate our algorithm by extensive experiments on synthetic datasets, on the standard UCI benchmark, and on speech separation.
Open AccessAnnealed Multiple Choice Learning: Overcoming limitations of Winner-takes-all with annealing
We introduce Annealed Multiple Choice Learning (aMCL) which combines simulated annealing with MCL. MCL is a learning framework handling ambiguous tasks by predicting a small set of plausible hypotheses. These hypotheses are trained using the Winner-takes-all (WTA) scheme, which promotes the diversity of the predictions. However, this scheme may converge toward an arbitrarily suboptimal local minimum, due to the greedy nature of WTA. We overcome this limitation using annealing, which enhances the exploration of the hypothesis space during training. We leverage insights from statistical physics and information theory to provide a detailed description of the model training trajectory. Additionally, we validate our algorithm by extensive experiments on synthetic datasets, on the standard UCI benchmark, and on speech separation.