TensorFlow implementation of Deep Mutual Learning accepted by CVPR 2018.
Deep mutual learning provides a simple but effective way to improve the generalisation ability of a network by training collaboratively with a cohort of other networks.
- TensorFlow 1.3.1
- CUDA 8.0 and cuDNN 6.0
- Matlab
-
Please download the Market-1501 Dataset
-
Convert the image data into TFRecords
sh scripts/format_and_convert_market.sh
- Train MobileNets with DML
sh scripts/train_dml_mobilenet_on_market.sh
- Train MobileNet independently
sh scripts/train_ind_mobilenet_on_market.sh
- Extract features of the test image
sh scripts/evaludate_dml_mobilenet_on_market.sh
- Evaluate the performance with matlab code
If you find DML useful in your research, please kindly cite our paper:
@inproceedings{ying2018DML,
author = {Ying Zhang and Tao Xiang and Timothy M. Hospedales and Huchuan Lu},
title = {Deep Mutual Learning},
booktitle = {CVPR},
year = {2018}}