Compression and Speed-up of Convolutional Neural Networks Through Dimensionality Reduction for Efficient Inference on Embedded Multiprocessor

L. Fernandez-Brillet (1, 2) , N. Leclaire (1, 2) , Stéphane Mancini (2) , M. Nicolas (1) , S. Cleyet-Merle (1) , J.P. Henriques (1) , C. Delnondedieu (1)

1 ST-GRENOBLE - STMicroelectronics [Grenoble]
2 TIMA-CDSI - Circuits, Devices and System Integration

Domaines: Traitement du signal et de l'image [eess.SP]

L. Fernandez-Brillet, N. Leclaire, Stéphane Mancini, M. Nicolas, S. Cleyet-Merle, et al.. Compression and Speed-up of Convolutional Neural Networks Through Dimensionality Reduction for Efficient Inference on Embedded Multiprocessor. Journal of Signal Processing Systems, 2021, ⟨10.1007/s11265-020-01616-0⟩. ⟨hal-03115158⟩