Romain Serizel
Présentation
I was born in Lille, France. I received the M.Eng. degree in Automatic System Engineering from ENSEM (Nancy, France) in 2005 and the M.Sc. degree in Signal Processing from Université Rennes 1 (Rennes, France) in 2006. I received the Ph.D. degree in Engineering Sciences from the Katholieke Universiteit Leuven (KUL), Belgium in June 2011. From June 2011 to December 2012 I was a research assistant with Prof. Marc Moonen research group at the Electrical Engineering Department (ESAT-SCD) of the KUL. From January 2013 to September 2014 i was a postdoctoral researcher with the Human Language Technology (HLT) group at the Fondazione Bruno Kessler (FBK) in Trento (Italy). From October 2014 to August 2016 I was a postdoctoral researcher with the department of signal and image processing (TSI) at telecom ParisTech in Paris (France) where I am working on speaker diarization, speaker recognition and respresentation learning for audio in the framework of the LASIE project. I am now an assistant professor at Université de Lorraine doing my research in Laboratoire lorrain de recherche en informatique et ses applications (Loria) with the Multispeech research group and teaching at IUT Charlemagne.
Publications
Publications
|
|
Diffused Responsibility: Analyzing the Energy Consumption of Generative Text-to-Audio Diffusion ModelsIEEE Workshop on Applications of Signal Processing to Audio and Acoustics, Oct 2025, Granlibakken Tahoe, United States |
|
|
Speaker Embeddings to Improve Tracking of Intermittent and Moving Speakers33rd European Signal Processing Conference (EUSIPCO 2025), Sep 2025, Palerme (Italie), Italy |
|
|
Diffusion-based Unsupervised Audio-visual Speech EnhancementICASSP 2025 - International Conference on Acoustics Speech and Signal Processing, IEEE, Apr 2025, Hyderabad, India. pp.1-5 |
|
|
A decade of DCASE: Achievements, practices, evaluations and future challengesICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Apr 2025, Hyderabad, France. pp.1-5, ⟨10.1109/ICASSP49660.2025.10887673⟩ |
|
|
Towards Low-Latency Tracking of Multiple Speakers With Short-Context Speaker Embeddings2025 IEEE 27th International Workshop on Multimedia Signal Processing (MMSP), Sep 2025, Beijin, China |
|
|
Evaluating Multichannel Speech Enhancement Algorithms at the Phoneme Scale Across Genders33rd European Signal Processing Conference, Sep 2025, Palerme, Italy. ⟨10.48550/arXiv.2506.18691⟩ |
|
|
Frequency-Weighted Training Losses for Phoneme-Level DNN-based Speech EnhancementIEEE 27th International Workshop on Multimedia Signal Processing, Sep 2025, Beijing, China. ⟨10.48550/arXiv.2506.18714⟩ |
|
|
Phoneme-Level Speech Intelligibility ReductionEUSIPCO 2025 - 33rd European Signal Processing Conference, Sep 2025, Palerme, Italy |
|
|
Tracking of Intermittent and Moving Speakers : Dataset and MetricsProceedings of the 11th Convention of the European Acoustics Association Forum Acusticum 2025, Jun 2025, Malaga, Espagne, Spain |
|
|
Unsupervised speech enhancement with diffusion-based generative modelsInternational Conference on Acoustics Speech and Signal Processing (ICASSP), IEEE, Apr 2024, Seoul (Korea), South Korea. ⟨10.48550/arXiv.2309.10450⟩ |
|
|
Normalizing Energy Consumption for Hardware-Independent Evaluation2024 IEEE International Workshop on Machine Learning for Signal Processing, Sep 2024, London, United Kingdom |
|
|
A weighted-variance variational autoencoder model for speech enhancementICASSP 2024 - International Conference on Acoustics Speech and Signal Processing, IEEE, Apr 2024, Seoul (Korea), South Korea. pp.1-5, ⟨10.1109/ICASSP48485.2024.10446294⟩ |
|
|
Mixture of Mixups for Multi-label Classification of Rare Anuran SoundsEUSIPCO 2024: 32nd European Signal Processing Conference, Aug 2024, Lyon, France. pp.1282-1286, ⟨10.23919/EUSIPCO63174.2024.10715140⟩ |
|
|
Posterior sampling algorithms for unsupervised speech enhancement with recurrent variational autoencoderInternational Conference on Acoustics Speech and Signal Processing (ICASSP), IEEE, Apr 2024, Seoul (Korea), South Korea. ⟨10.48550/arXiv.2309.10439⟩ |
|
|
Self-Supervised Learning for Few-Shot Bird Sound ClassificationICASSPW 2024: IEEE International Conference on Acoustics, Speech, and Signal Processing Workshops, Apr 2024, Seoul, South Korea. ⟨10.1109/ICASSPW62465.2024.10627576⟩ |
|
|
RoboVox: A Single/Multi-channel Far-field Speaker Recognition Benchmark for a Mobile RobotLREC_COLING, ELRA, May 2024, Turino, Italy |
|
|
From Computation to Consumption: Exploring the Compute-Energy Link for Training and Testing Neural Networks for SED SystemsDetection and Classification of Acoustic Scenes and Events 2024, Oct 2024, Tokyo, Japan |
|
|
Regularized Contrastive Pre-Training for Few-Shot Bioacoustic Sound DetectionICASSP 2024: IEEE International Conference on Acoustics, Speech and Signal Processing, Apr 2024, Seoul, South Korea. ⟨10.1109/ICASSP48485.2024.10446409⟩ |
|
|
Performance and Energy Balance: A Comprehensive Study of State-of-the-Art Sound Event Detection SystemsICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Apr 2024, Seoul, South Korea. pp.1096-1100, ⟨10.1109/ICASSP48485.2024.10445834⟩ |
|
|
Diffusion-based speech enhancement with a weighted generative-supervised learning lossInternational Conference on Acoustics Speech and Signal Processing (ICASSP), IEEE, Apr 2024, Seoul (Korea), South Korea. ⟨10.48550/arXiv.2309.10457⟩ |
|
|
Multi-channel extension of pre-trained models for speaker verificationInterspeech, Sep 2024, Kos, Greece |
Monitoring environmental impact of DCASE systems: Why and how ?Detection and Classification of Acoustic Scene and Events (DCASE) Workshop, Sep 2023, Tampere (Finlande), Finland |
|
|
|
Performance above all ? energy consumption vs. performance for machine listening, a study on dcase task 4 baselineICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Jun 2023, Rhodes Island, France. pp.1-5, ⟨10.1109/ICASSP49357.2023.10095938⟩ |
|
|
Post-Processing Independent Evaluation of Sound Event Detection SystemsDCASE 2023 - 8th Workshop on Detection and Classification of Acoustic Scenes and Events, Sep 2023, Tampere, Finland. ⟨10.48550/arXiv.2306.15440⟩ |
|
|
Self-supervised learning with diffusion-based multichannel speech enhancement for speaker verification under noisy conditionsINTERSPEECH 2023, Aug 2023, Dublin (Ireland), Ireland. pp.3849-3853, ⟨10.21437/Interspeech.2023-1890⟩ |
|
|
Pretraining Representations for Bioacoustic Few-Shot Detection using Supervised Contrastive LearningDetection and Classification of Acoustic Scenes and Events 2023, Sep 2023, TAMPERE, Finland |
|
|
Fast and efficient speech enhancement with variational autoencodersInternational Conference on Acoustics Speech and Signal Processing (ICASSP), IEEE, Jun 2023, Rhodes island, Greece |
|
|
Joint optimization of diffusion probabilistic-based multichannel speech enhancement with far-field speaker verificationIEEE SLT 2022, Jan 2023, Doha, Qatar |
|
|
Lightweight Annotation and Class Weight Training for Automatic Estimation of Alarm Audibility in NoiseICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing, Jun 2023, Rhodes Island, Greece. pp.1-5, ⟨10.1109/ICASSP49357.2023.10094730⟩ |
|
|
SPICE+: Evaluation of automatic audio captioning systems with pre-trained language modelsIEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2023), Jun 2023, Rhodes Island, Greece |
|
|
BinauRec: A dataset to test the influence of the use of room impulse responses on binaural speech enhancementEUSIPCO23, EURASIP, Sep 2023, Helsiinki, Finland. ⟨10.23919/EUSIPCO58844.2023.10289772⟩ |
|
|
From Discrete Tokens to High-Fidelity Audio Using Multi-Band DiffusionNeurIPS 2023 - Conference on Neural Information Processing Systems, Dec 2023, New Orleans, United States. ⟨10.48550/arXiv.2308.02560⟩ |
|
|
Audio-visual speech enhancement with a deep kalman filter generative modelInternational Conference on Acoustics Speech and Signal Processing (ICASSP), IEEE, Jun 2023, Rhodes island, Greece |
|
|
CONVOLUTIONAL NEURAL NETWORK FOR AUDIBILITY ASSESSMENT OF ACOUSTIC ALARMSInternational Conference on Acoustics, Speech and Signal Processing Search form Search ICASSP IEEE 2023, Jun 2023, Rhodes, Greece |
|
|
Learning noise robust ResNet-based speaker embedding for speaker recognitionOdyssey 2022 : The Speaker and Language Recognition Workshop, Jun 2022, Beijing, China |
|
|
Threshold independent evaluation of sound event detection scoresICASSP 2022 - IEEE International Conference on Acoustics, Speech and Signal Processing, May 2022, Singapore, Singapore. ⟨10.1109/ICASSP43922.2022.9747556⟩ |
|
|
Barlow Twins self-supervised learning for robust speaker recognitionInterspeech 2022 - Human and Humanizing Speech Technology, Sep 2022, Incheon, South Korea. ⟨10.21437/Interspeech.2022-11301⟩ |
|
|
Integrating isolated examples with weakly-supervised sound event detection: a direct approach7th Workshop on Detection and Classification of Acoustic Scenes and Events (DCASE), Nov 2022, Nancy, France |
Evaluation de l’audibilité ressentie des alarmes sonores dans le bruit16ème Congrès Français d'Acoustique, CFA2022, Société Française d'Acoustique; Laboratoire de Mécanique et d'Acoustique, Apr 2022, Marseille, France |
|
|
|
A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapesICASSP 2022 - IEEE International Conference on Acoustics, Speech and Signal Processing, May 2022, Singapore/Virtual, Singapore. ⟨10.1109/ICASSP43922.2022.9747577⟩ |
|
|
How to Leverage DNN-based speech enhancement for multi-channel speaker verification?4th International Conference on Advances in Signal Processing and Artificial Intelligence (ASPAI' 2022), Oct 2022, Corfu, Greece |
Modélisation de la détection d’alarmes sonores dans le bruit16ème Congrès Français d'Acoustique, CFA2022, Société Française d'Acoustique; Laboratoire de Mécanique et d'Acoustique, Apr 2022, Marseille, France |
|
|
|
A Comprehensive Exploration of Noise Robustness and Noise Compensation in ResNet and TDNN-based Speaker Recognition SystemsEUSIPCO 2022 - 30th European Signal Processing Conference, Aug 2022, Belgrade, Serbia |
|
|
Foreground-Background Ambient Sound Scene SeparationEUSIPCO 2020 - 28th European Signal Processing Conference, Jan 2021, Amsterdam / Virtual, Netherlands. ⟨10.23919/Eusipco47968.2020.9287436⟩ |
|
|
UIAI System for Short-Duration Speaker Verification Challenge 2020SLT 2021 - IEEE Spoken Language Technology Workshop, IEEE, Jan 2021, Shenzhen / Virtual, China. ⟨10.1109/SLT48900.2021.9383596⟩ |
|
|
What’s All the FUSS About Free Universal Sound Separation Data?ICASSP 2021 - 46th International Conference on Acoustics, Speech, and Signal Processing, Jun 2021, Toronto/Virtual, Canada. ⟨10.1109/ICASSP39728.2021.9414774⟩ |
|
|
Automated audio captioning by fine-tuning bart with audioset tagsDCASE 2021 - 6th Workshop on Detection and Classification of Acoustic Scenes and Events, Nov 2021, Virtual, Spain |
|
|
Attention-based distributed speech enhancement for unconstrained microphone arrays with varying number of nodesEUSIPCO 2021 - 29th European Signal Processing Conference, IEEE, Aug 2021, Dublin / Virtual, Ireland. ⟨10.23919/EUSIPCO54536.2021.9616358⟩ |
|
|
Sound Event Detection and Separation: a Benchmark on Desed Synthetic SoundscapesICASSP 2021 - 46th International Conference on Acoustics, Speech, and Signal Processing, Jun 2021, Toronto/Virtual, Canada. ⟨10.1109/ICASSP39728.2021.9414789⟩ |
|
|
Distributed speech separation in spatially unconstrained microphone arraysICASSP 2021 - 46th International Conference on Acoustics, Speech, and Signal Processing, Jun 2021, Toronto / Virtual, Canada. ⟨10.1109/ICASSP39728.2021.9414758⟩ |
|
|
Improving Sound Event Detection Metrics: Insights from DCASE 2020ICASSP 2021 - 46th International Conference on Acoustics, Speech, and Signal Processing, Jun 2021, Toronto/Virtual, Canada. ⟨10.1109/ICASSP39728.2021.9414711⟩ |
|
|
The impact of non-target events in synthetic soundscapes for sound event detectionDCASE 2021 - Detection and Classification of Acoustic Scenes and Events, Nov 2021, Barcelona/Virtual, Spain |
|
|
Compensate multiple distortions for speaker recognition systemsEUSIPCO 2021 - 29th European Signal Processing Conference, Aug 2021, Dublin / Virtual, Ireland. ⟨10.23919/EUSIPCO54536.2021.9615983⟩ |
|
|
A brief introduction to multichannel noise reduction with deep neural networksSpiN 2020 - 12th Speech in Noise Workshop, Jan 2020, Toulouse, France |
|
|
Improving Sound Event Detection In Domestic Environments Using Sound SeparationDCASE Workshop 2020 - Detection and Classification of Acoustic Scenes and Events, Nov 2020, Tokyo / Virtual, Japan |
|
|
DNN-Based Distributed Multichannel Mask Estimation for Speech Enhancement in Microphone ArraysICASSP 2020 - 45th International Conference on Acoustics, Speech, and Signal Processing, May 2020, Barcelona, Spain |
|
|
Training Sound Event Detection On A Heterogeneous DatasetDCASE Workshop, Nov 2020, Tokyo, Japan |
|
|
Sound event detection in synthetic domestic environmentsICASSP 2020 - 45th International Conference on Acoustics, Speech, and Signal Processing, May 2020, Barcelona, Spain |
|
|
Limitations of weak labels for embedding and taggingICASSP 2020 - 45th International Conference on Acoustics, Speech, and Signal Processing, May 2020, Barcelona, Spain |
|
|
Semi-supervised triplet loss based learning of ambient audio embeddingsICASSP 2019, May 2019, Brighton, United Kingdom |
|
|
Sound event detection in domestic environments with weakly labeled data and soundscape synthesisWorkshop on Detection and Classification of Acoustic Scenes and Events, Oct 2019, New York City, United States |
|
|
Regression versus classification for neural network based audio source localizationWASPAA 2019 - IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, IEEE, Oct 2019, New Paltz, United States |
|
|
Sound Event Detection from Partially Annotated Data: Trends and ChallengesIcETRAN conference, Jun 2019, Srebrno Jezero, Serbia |
|
|
Multiple-input neural network-based residual echo suppressionICASSP 2018 - IEEE International Conference on Acoustics, Speech and Signal Processing, Apr 2018, Calgary, Canada. pp.1-5 |
|
|
Large-Scale Weakly Labeled Semi-Supervised Sound Event Detection in Domestic EnvironmentsWorkshop on Detection and Classification of Acoustic Scenes and Events, Nov 2018, Woking, United Kingdom |
|
|
Multichannel speech separation with recurrent neural networks from high-order ambisonics recordings43rd IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2018), Apr 2018, Calgary, Canada |
|
|
Multichannel Audio Modeling with Elliptically Stable Tensor DecompositionLVA/ICA 2018 - 14th International Conference on Latent Variable Analysis and Signal Separation, Jul 2018, Surrey, United Kingdom. pp.13-23, ⟨10.1007/978-3-319-93764-9_2⟩ |
|
|
CRNN-based joint azimuth and elevation localization with the Ambisonics intensity vectorIWAENC 2018 - 16th International Workshop on Acoustic Signal Enhancement, Sep 2018, Tokyo, Japan |
|
|
Leveraging deep neural networks with nonnegative representations for improved environmental sound classificationIEEE International Workshop on Machine Learning for Signal Processing MLSP, Sep 2017, Tokyo, Japan |
|
|
Supervised Group Nonnegative Matrix Factorisation With Similarity Constraints And Applications To Speaker IdentificationIEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Mar 2017, New Orleans, United States |
|
|
Nonnegative Feature Learning Methods for Acoustic Scene ClassificationDCASE 2017 - Workshop on Detection and Classification of Acoustic Scenes and Events, Nov 2017, Munich, Germany |
Acoustic scene classification with matrix factorization for unsupervised feature learningICASSP, Mar 2016, Shangai, China |
|
|
|
Machine listening techniques as a complement to video image analysis in forensicsIEEE International Conference on Image Processing, Sep 2016, Phoenix, AZ, United States. pp.948-952, ⟨10.1109/ICIP.2016.7532497⟩ |
|
|
SUPERVISED NONNEGATIVE MATRIX FACTORIZATION FOR ACOUSTIC SCENE CLASSIFICATIONIEEE international evaluation campaign on detection and classification of acousitc scenes and events (DCASE 2016), Sep 2016, Budapest, Hungary |
Group nonnegative matrix factorisation with speaker and session variability compensation for speaker identificationICASSP, Mar 2016, Shangai, China. pp.5470 - 5474 |
|
|
|
Group Non-Negative Matrix Factorisation With Speaker And Session Similarity Constraints For Speaker IdentificationIEEE International Conference on Acoustics, Speech, and Signal Processing, Mar 2016, Shangai, China |
|
|
Mini-batch stochastic approaches for accelerated multiplicative updates in nonnegative matrix factorisation with beta-divergenceIEEE International Workshop on Machine Learning for Signal Processing (MLSP 2016), Sep 2016, Salerne, Italy |
A brief introduction to deep neural networks and their application to automatic speech recognitionSéminaire de l'équipe Perception, Feb 2015, Grenoble, France |
|
|
|
Vocal tract length normalisation approaches to DNN-based children's and adults' speech recognition2014 IEEE Spoken Language Technology Workshop (SLT 2014), Dec 2014, South Lake Tahoe, CA, United States. pp.135-140, ⟨10.1109/SLT.2014.7078563⟩ |
|
|
Deep neural network adaptation for children's and adults' speech recognitionItalian Computational Linguistics Conference (CLiC-it), Dec 2014, Pise, Italy |
|
|
Rank-1 Approximation Based Multichannel Wiener Filtering Algorithms For Noise Reduction In Cochlear ImplantsIEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), May 2013, Vancouver, Canada |
|
|
Output SNR analysis of integrated active noise control and noise reduction in hearing aids under a single speech source scenarioEuropean Signal Processing Conference (EUSIPCO), Aug 2010, Aalborg, Denmark |
|
|
A Zone of Quiet Based Approach to Integrated Active Noise Control and Noise Reduction in Hearing AidsIEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), Oct 2009, New Paltz, United States |
|
|
A Weighted Approach for Integrated Active Noise Control and Noise Reduction in Hearing AidsEuropean Signal Processing Conference (EUSIPCO), Aug 2009, Glasgow, United Kingdom |
|
|
Combined Active Noise Control and noise reduction in Hearing AidsInternational Workshop on Acoustic Echo and Noise Control (IWAENC), Sep 2008, Seattle, United States |
Noise model for Accuracy Constraint Determination in Fixed-Point SystemsWorkshop on Design and Architectures for Signal and Image Processing DASIP 2007, Nov 2007, Grenoble, France |
|
|
Multiview approaches to event detection and scene analysisTuomas Virtanen; Mark D. Plumbley; Dan Ellis. Computational Analysis of Sound Scenes and Events, Springer, pp.243-276, 2017, 978-3319634494. ⟨10.1007/978-3-319-63450-0_9⟩ |
|
|
Acoustic Features for Environmental Sound AnalysisTuomas Virtanen; Mark D. Plumbley; Dan Ellis. Computational Analysis of Sound Scenes and Events, Springer International Publishing AG, pp.71-101, 2017, 978-3-319-63449-4. ⟨10.1007/978-3-319-63450-0_4⟩ |
|
|
Supervised contrastive learning for pre-training bioacoustic few-shot systemsIMT Atlantique; LORIA. 2023 |
|
|
Robovox: Far-Field Speaker Recognition By A Mobile Robot (Evaluation Plan)Avignon Université. 2023 |
|
|
Joint NN-Supported Multichannel Reduction of Acoustic Echo, Reverberation and Noise: Supporting Document[Research Report] RR-9303, INRIA Nancy; Invoxia SAS. 2019 |
|
|
Contributions to speech processing and ambient sound analysisComputer Science [cs]. Université de Lorraine, 2022 |