Deep Model Pruning without Finetuning for Few Category Datasets

Current deep model pruning methods mainly focus on large datasets and often involve finetuning before deployment. However, in real-world applications, pruning is typically needed for open scenarios with few classification categories, where fine-tuning should be avoided to preserve the generalization of deep models. To address these issues, we propose a novel deep model pruning method called Cluster-based Redundancy Elimination (CRE). Specifically, CRE first represents each convolutional kernel as a point in high-dimensional space. Second, a distance-based strategy is employed to compute a clustering radius for each convolutional layer. Finally, based on these clustering radii, core point filters are selected for pruning, as they extract redundant information that can be captured by neighboring filters in the high-dimensional space. Thus, finetuning deep models to recover their performance can be removed. Comprehensive experiments on five datasets with few categories validate the effectiveness of our approach and demonstrate its superiority over several state-of-the-art pruning methods.

I. Introduction

Convolutional neural networks (CNNs) have demonstrated exceptional success in various computer vision tasks, such as image classification [1], object detection [2], and semantic segmentation [3]. Generally, deeper and wider CNN architectures lead to better performance [4]. However, the significant computational costs associated with such architectures limit their deployment on resource-constrained devices, such as mobile and edge platforms. As a result, pruning deep neural networks to reduce their storage and computational demands has become an active area of research [5] –[9].