%0 Conference Proceedings
%T Energy-Accuracy Scalable Deep Convolutional Neural Networks: A Pareto Analysis
%+ Department of Computer Engineering (DAUIN)
%A Peluso, Valentino
%A Calimera, Andrea
%< avec comité de lecture
%( IFIP Advances in Information and Communication Technology
%B 26th IFIP/IEEE International Conference on Very Large Scale Integration - System on a Chip (VLSI-SoC)
%C Verona, Italy
%Y Nicola Bombieri
%Y Graziano Pravadelli
%Y Masahiro Fujita
%Y Todd Austin
%Y Ricardo Reis
%I Springer International Publishing
%3 VLSI-SoC: Design and Engineering of Electronics Systems Based on New Computing Paradigms
%V AICT-561
%P 107-127
%8 2018-10-08
%D 2018
%R 10.1007/978-3-030-23425-6_6
%Z Computer Science [cs]Conference papers
%X This work deals with the optimization of Deep Convolutional Neural Networks (ConvNets). It elaborates on the concept of Adaptive Energy-Accuracy Scaling through multi-precision arithmetic, a solution that allows ConvNets to be adapted at run-time and meet different energy budgets and accuracy constraints. The strategy is particularly suited for embedded applications made run at the “edge” on resource-constrained platforms. After the very basics that distinguish the proposed adaptive strategy, the paper recalls the software-to-hardware vertical implementation of precision scalable arithmetic for ConvNets, then it focuses on the energy-driven per-layer precision assignment problem describing a meta-heuristic that searches for the most suited representation of both weights and activations of the neural network. The same heuristic is then used to explore the optimal trade-off providing the Pareto points in the energy-accuracy space. Experiments conducted on three different ConvNets deployed in real-life applications, i.e. Image Classification, Keyword Spotting, and Facial Expression Recognition, show adaptive ConvNets reach better energy-accuracy trade-off w.r.t. conventional static fixed-point quantization methods.
%G English
%Z TC 10
%Z WG 10.5
%2 https://inria.hal.science/hal-02321763/document
%2 https://inria.hal.science/hal-02321763/file/485996_1_En_6_Chapter.pdf
%L hal-02321763
%U https://inria.hal.science/hal-02321763
%~ IFIP
%~ IFIP-AICT
%~ IFIP-TC
%~ IFIP-WG
%~ IFIP-VLSISOC
%~ IFIP-TC10
%~ IFIP-WG10-5
%~ IFIP-AICT-561