GAAF: Searching Activation Functions for Binary Neural Networks through Genetic Algorithm

TL;DR

This paper introduces GAAF, a genetic algorithm-based method to automatically search for effective activation functions in binary neural networks, significantly improving their accuracy on various datasets.

Contribution

Proposes a novel GA-based search for activation functions in BNNs, leading to 15 new functions that enhance performance and gradient approximation.

Findings

Achieved up to 2.54% accuracy improvement on ImageNet

Identified 15 novel activation functions through GA search

Enhanced BNN performance across multiple datasets and models

Abstract

Binary neural networks (BNNs) show promising utilization in cost and power-restricted domains such as edge devices and mobile systems. This is due to its significantly less computation and storage demand, but at the cost of degraded performance. To close the accuracy gap, in this paper we propose to add a complementary activation function (AF) ahead of the sign based binarization, and rely on the genetic algorithm (GA) to automatically search for the ideal AFs. These AFs can help extract extra information from the input data in the forward pass, while allowing improved gradient approximation in the backward pass. Fifteen novel AFs are identified through our GA-based search, while most of them show improved performance (up to 2.54% on ImageNet) when testing on different datasets and network models. Our method offers a novel approach for designing general and application-specific BNN architecture. Our code is available at http://github.com/flying-Yan/GAAF.