One-shot Network Pruning at Initialization with Discriminative Image Patches
This work addresses the efficiency of neural network pruning for machine learning practitioners, but it is incremental as it builds on existing OPaI methods with data-dependent improvements.
The paper tackled the problem of one-shot network pruning at initialization by showing that data is crucial for performance, contrary to recent beliefs, and introduced Discriminative One-shot Pruning and Super Stitching methods that significantly outperform original methods on ImageNet, especially in highly compressed models.
One-shot Network Pruning at Initialization (OPaI) is an effective method to decrease network pruning costs. Recently, there is a growing belief that data is unnecessary in OPaI. However, we obtain an opposite conclusion by ablation experiments in two representative OPaI methods, SNIP and GraSP. Specifically, we find that informative data is crucial to enhancing pruning performance. In this paper, we propose two novel methods, Discriminative One-shot Network Pruning (DOP) and Super Stitching, to prune the network by high-level visual discriminative image patches. Our contributions are as follows. (1) Extensive experiments reveal that OPaI is data-dependent. (2) Super Stitching performs significantly better than the original OPaI method on benchmark ImageNet, especially in a highly compressed model.