Backbone Can Not be Trained at Once: Rolling Back to Pre-trained Network for Person Re-Identification
arXiv:1901.06140
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.
arXiv commentAccepted to AAAI 2019
References in corpus (13)
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
- Distilling the Knowledge in a Neural Network
- In Defense of the Triplet Loss for Person Re-Identification
- SGDR: Stochastic Gradient Descent with Warm Restarts
- Fine-Grained Visual Classification of Aircraft
- Person Re-identification: Past, Present and Future
- Random Erasing Data Augmentation
- Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitro
- Deeply-Learned Part-Aligned Representations for Person Re-Identification
- SVDNet for Pedestrian Retrieval
- DeeperCut: A Deeper, Stronger, and Faster Multi-Person Pose Estimation Model
- Attention-Aware Compositional Network for Person Re-identification
- Deep Mutual Learning