The Costs of Reproducibility in Music Separation Research: a Replication of Band-Split RNN.

Constance Douwes

Paul Magron Romain Serizel Constance Douwes The Costs of Reproducibility in Music Separation Research: a Replication of Band-Split RNN. 2026 March abs/2603.09187 CoRR https://doi.org/10.48550/arXiv.2603.09187 db/journals/corr/corr2603.html#abs-2603-09187 streams/journals/corr

Constance Douwes Romain Serizel Energy Consumption Trends in Sound Event Detection Systems. 1-5 2025 ICASSP https://doi.org/10.1109/ICASSP49660.2025.10889562 conf/icassp/2025 db/conf/icassp/icassp2025.html#DouwesS25 Constance Douwes Romain Serizel Normalizing Energy Consumption for Hardware-Independent Evaluation. 1-6 2024 MLSP https://doi.org/10.1109/MLSP58920.2024.10734733 conf/mlsp/2024 db/conf/mlsp/mlsp2024.html#DouwesS24 streams/conf/mlsp

Samuele Cornell Janek Ebbers Constance Douwes Irene Martín-Morató Manu Harju Annamaria Mesaros Romain Serizel DCASE 2024 Task 4: Sound Event Detection with Heterogeneous Data and Missing Labels. 2024 abs/2406.08056 CoRR https://doi.org/10.48550/arXiv.2406.08056 db/journals/corr/corr2406.html#abs-2406-08056 streams/journals/corr

Constance Douwes Romain Serizel From Computation to Consumption: Exploring the Compute-Energy Link for Training and Testing Neural Networks for SED Systems. 2024 abs/2409.05080 CoRR https://doi.org/10.48550/arXiv.2409.05080 db/journals/corr/corr2409.html#abs-2409-05080 streams/journals/corr

Constance Douwes Romain Serizel Normalizing Energy Consumption for Hardware-Independent Evaluation. 2024 abs/2409.05602 CoRR https://doi.org/10.48550/arXiv.2409.05602 db/journals/corr/corr2409.html#abs-2409-05602 streams/journals/corr

Constance Douwes Romain Serizel Energy Consumption Trends in Sound Event Detection Systems. 2024 abs/2409.08763 CoRR https://doi.org/10.48550/arXiv.2409.08763 db/journals/corr/corr2409.html#abs-2409-08763 streams/journals/corr

Constance Douwes Hierarchical variational temporal learning for dynamic musical audio synthesis. (Sur l'impact environnemental des modèles génératifs profonds pour l'audio). 2023 Sorbonne University, Paris, France https://tel.archives-ouvertes.fr/tel-04100243 Constance Douwes Giovanni Bindi Antoine Caillon Philippe Esling Jean-Pierre Briot Is Quality Enoughƒ Integrating Energy Consumption in a Large-Scale Evaluation of Neural Audio Synthesis Models. 1-5 2023 ICASSP https://doi.org/10.1109/ICASSP49357.2023.10096975 conf/icassp/2023 db/conf/icassp/icassp2023.html#DouwesBCEB23

Constance Douwes Philippe Esling Jean-Pierre Briot A Multi-Objective Approach for Sustainable Generative Audio Models. 2021 abs/2107.02621 CoRR https://arxiv.org/abs/2107.02621 db/journals/corr/corr2107.html#abs-2107-02621

Philippe Esling Ninon Devis Adrien Bitton Antoine Caillon Axel Chemla-Romeu-Santos Constance Douwes Diet deep generative audio models with structured lottery. 2020 abs/2007.16170 CoRR https://arxiv.org/abs/2007.16170 db/journals/corr/corr2007.html#abs-2007-16170

Giovanni Bindi Adrien Bitton Jean-Pierre Briot Antoine Caillon Axel Chemla-Romeu-Santos Samuele Cornell Ninon Devis Janek Ebbers Philippe Esling Manu Harju Paul Magron Irene Martín-Morató Annamaria Mesaros Romain Serizel