Efficient AI for Sound using Pruning

Our poster titled "Efficient AI for Sound using Pruning" has been awarded the Best Poster Award at the Third UK AI Conference (UKAI 2025) 🎉 The work focuses on designing lightweight and energy-efficient neural networks for audio processing—an important step toward sustainable AI. Honoured to have our efforts recognised among so many inspiring contributions in the UK AI research community. This work is a part of EPSRC funded AI for sound (AI4S) project at Centre for Vision, Speech and Signal Processing at University of Surrey Surrey Institute for People-Centred AI (PAI)is a summary of following collaborative work: ----- A Singh and M. D. Plumbley, "Efficient CNNs via Passive Filter Pruning," in IEEE Transactions on Audio, Speech and Language Processing, doi: 10.1109/TASLPRO.2025.3561589. -----X. Xu, A. Singh, M. Wu, W. Wang and M. D. Plumbley, "Investigating Passive Filter Pruning for Efficient CNN-Transformer Audio Captioning," IEEE 34th International Workshop on Machine Learning for Signal Processing (MLSP), doi: 10.1109/MLSP58920.2024.10734745, 2024. -----Gabriel Bibbo, A Singh, Mark D. Plumbley, "Environmental sound classification on an embedded hardware platform", INTER-NOISE and NOISE-CON Congress and Conference Proceedings, INTER-NOISE24, Nantes, France, pp. 6376-6385, 2024. ----J. A. King, A. Singh and M. D. Plumbley, "Compressing Audio CNNS with Graph Centrality Based Filter Pruning," 2023 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), New Paltz, NY, USA, doi: 10.1109/WASPAA58266.2023.10248103, 2023. ----A Singh, Haohe Liu, and Mark D. Plumbley. "E-PANNs: sound recognition using efficient pre-trained audio neural networks." Inter-Noise and Noise-Con Congress and Conference Proceedings, Institute of Noise Control Engineering, vol 268 No 1, pp 7220-7228, 2023. UK AI Conference link: https://uk-ai.org/ukai2025/ AI for sound project: https://ai4s.surrey.ac.uk/