Backbone Can Not be Trained at Once: Rolling Back to Pre-trained Network for Person Re-Identification

TL;DR

This paper introduces a novel fine-tuning strategy for person re-identification that involves rolling back high-level layers to their pre-trained weights, effectively training low-level layers despite gradient vanishing issues, leading to state-of-the-art results.

Contribution

The paper proposes a new fine-tuning method that resets high-level layers to pre-trained weights to improve low-level layer training in ReID tasks, addressing gradient vanishing.

Findings

Enhanced ReID performance without additional modules

State-of-the-art results achieved with vanilla CNNs

Effective training of low-level layers via weight rollback

Abstract

In person re-identification (ReID) task, because of its shortage of trainable dataset, it is common to utilize fine-tuning method using a classification network pre-trained on a large dataset. However, it is relatively difficult to sufficiently fine-tune the low-level layers of the network due to the gradient vanishing problem. In this work, we propose a novel fine-tuning strategy that allows low-level layers to be sufficiently trained by rolling back the weights of high-level layers to their initial pre-trained weights. Our strategy alleviates the problem of gradient vanishing in low-level layers and robustly trains the low-level layers to fit the ReID dataset, thereby increasing the performance of ReID tasks. The improved performance of the proposed strategy is validated via several experiments. Furthermore, without any add-ons such as pose estimation or segmentation, our strategy exhibits state-of-the-art performance using only vanilla deep convolutional neural network architecture.