Backbone Can Not be Trained at Once: Rolling Back to Pre-trained Network for Person Re-Identification
Cover image
Training loss and mAP graph changed by introducing our learning strategy. `base' means that the network is trained by basic strategy. In our method, the training loss escapes from local minimum and the mAP accuracy increases by utilizing the rolling-back scheme.
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.
Publication
Backbone Can Not be Trained at Once: Rolling Back to Pre-trained Network for Person Re-Identification
Youngmin Ro, Jongwon Choi, Dae Ung Jo, Byeongho Heo, Jongin Lim, Jin Young Choi
33rd AAAI Conference on Artificial Intelligence (AAAI), 2019
If you have questions, please contact youngmin4920@gmail.com