CATRO: Channel Pruning via Class-Aware Trace Ratio Optimization

TL;DR

CATRO is a novel class-aware channel pruning method that optimizes feature space discriminations to produce lightweight neural networks with guaranteed performance improvements, suitable for various classification tasks.

Contribution

It introduces a class-aware trace ratio optimization framework for channel pruning, with theoretical convergence guarantees and superior empirical results.

Findings

Achieves higher accuracy with similar or lower computation cost.

Effectively prunes networks for various classification subtasks.

Demonstrates theoretical convergence and performance guarantees.

Abstract

Deep convolutional neural networks are shown to be overkill with high parametric and computational redundancy in many application scenarios, and an increasing number of works have explored model pruning to obtain lightweight and efficient networks. However, most existing pruning approaches are driven by empirical heuristic and rarely consider the joint impact of channels, leading to unguaranteed and suboptimal performance. In this paper, we propose a novel channel pruning method via Class-Aware Trace Ratio Optimization (CATRO) to reduce the computational burden and accelerate the model inference. Utilizing class information from a few samples, CATRO measures the joint impact of multiple channels by feature space discriminations and consolidates the layer-wise impact of preserved channels. By formulating channel pruning as a submodular set function maximization problem, CATRO solves it efficiently via a two-stage greedy iterative optimization procedure. More importantly, we present theoretical justifications on convergence of CATRO and performance of pruned networks. Experimental results demonstrate that CATRO achieves higher accuracy with similar computation cost or lower computation cost with similar accuracy than other state-of-the-art channel pruning algorithms. In addition, because of its class-aware property, CATRO is suitable to prune efficient networks adaptively for various classification subtasks, enhancing handy deployment and usage of deep networks in real-world applications.