Channel Gating Neural Networks

TL;DR

This paper proposes channel gating, a dynamic pruning method for CNNs that reduces computation by skipping less important channels at run-time, achieving significant FLOP reduction with minimal accuracy loss.

Contribution

It introduces a novel input-dependent channel gating scheme that enhances CNN efficiency and demonstrates hardware-friendly implementation and acceleration.

Findings

Achieves 2.7-8.0× FLOP reduction on CIFAR-10.

Reduces ResNet-18 compute cost by 2.6× on ImageNet.

Provides hardware accelerator design for channel gating networks.

Abstract

This paper introduces channel gating, a dynamic, fine-grained, and hardware-efficient pruning scheme to reduce the computation cost for convolutional neural networks (CNNs). Channel gating identifies regions in the features that contribute less to the classification result, and skips the computation on a subset of the input channels for these ineffective regions. Unlike static network pruning, channel gating optimizes CNN inference at run-time by exploiting input-specific characteristics, which allows substantially reducing the compute cost with almost no accuracy loss. We experimentally show that applying channel gating in state-of-the-art networks achieves 2.7-8.0$\times$ reduction in floating-point operations (FLOPs) and 2.0-4.4$\times$ reduction in off-chip memory accesses with a minimal accuracy loss on CIFAR-10. Combining our method with knowledge distillation reduces the compute cost of ResNet-18 by 2.6$\times$ without accuracy drop on ImageNet. We further demonstrate that channel gating can be realized in hardware efficiently. Our approach exhibits sparsity patterns that are well-suited to dense systolic arrays with minimal additional hardware. We have designed an accelerator for channel gating networks, which can be implemented using either FPGAs or ASICs. Running a quantized ResNet-18 model for ImageNet, our accelerator achieves an encouraging speedup of 2.4$\times$ on average, with a theoretical FLOP reduction of 2.8$\times$.