Optimizing The Consumption Of Spiking Neural Networks With Activity Regularization

Reducing energy consumption is a critical point for neural network models running on edge devices. In this regard, reducing the number of multiply-accumulate (MAC) operations of Deep Neural Networks (DNNs) running on edge hardware accelerators will reduce the energy consumption during inference. Spi...

Full description

Saved in:
Bibliographic Details
Published inICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) pp. 61 - 65
Main Authors Narduzzi, Simon, Bigdeli, Siavash A., Liu, Shih-Chii, Dunbar, L. Andrea
Format Conference Proceeding
LanguageEnglish
Published IEEE 23.05.2022
Subjects
Deep learning
Deep Neural Networks
Encoding
Energy consumption
Neural networks
Regularization
Signal processing
Sparsity
Spiking Neural Networks
Stability analysis
Training
Online AccessGet full text

Cover

More Information
Summary:Reducing energy consumption is a critical point for neural network models running on edge devices. In this regard, reducing the number of multiply-accumulate (MAC) operations of Deep Neural Networks (DNNs) running on edge hardware accelerators will reduce the energy consumption during inference. Spiking Neural Networks (SNNs) are an example of bio-inspired techniques that can further save energy by using binary activations, and avoid consuming energy when not spiking. The networks can be configured for equivalent accuracy on a task through DNN-to-SNN conversion frameworks but their conversion is based on rate coding therefore the synaptic operations can be high. In this work, we look into different techniques to enforce sparsity on the neural network activation maps and compare the effect of different training regularizers on the efficiency of the optimized DNNs and SNNs.
ISSN:2379-190X
DOI:10.1109/ICASSP43922.2022.9746375