Vehicle_ReID_using_multi_feature

This is the source code for my Graduation Thesis: Vehicle Re-Identification using fusion of multi-features

System requirement

• Python - 2.7.14
• cuda - 8.0
• cudatoolkit - 7.5
• cudnn - 5.1
• tensorflow (-gpu) - 1.0.1
• keras (-gpu) - 2.1.5
• h5py - 2.7.1
• caffe - 1.0.0
• MatLab - R2017b

Addittional Sources

• VeRi-776 Dataset: https://github.com/VehicleReId/VeRidataset
• CompCars Dataset: http://mmlab.ie.cuhk.edu.hk/datasets/comp_cars/index.html
• Vcolor models: https://github.com/rezafuad/vehicle-color-recognition
• GoogleNet models: https://gist.github.com/bogger/b90eb88e31cd745525ae
• Trained models, results and my thesis are available at: https://drive.google.com/open?id=1NhSOgZlJqtUkjns2fva6nfNz18CJnXqH
• 'How to run' is available at: https://www.youtube.com/watch?v=gDNMbp4e1z8
• Spotlight Video is available at: https://www.youtube.com/watch?v=y_hsnMtaKQk

Author

• Hoang Trinh Man - hoangtrinh1001@gmail.com

Citation

• Liu X., Liu W., Ma H., Fu H.: Large-scale vehicle re-identification in urban surveillance videos. In: IEEE International Conference on Multimedia and Expo. (2016) accepted.
• Liu X., Liu W., Mei T., Ma H. A Deep Learning-Based Approach to Progressive Vehicle Re-identification for Urban Surveillance. In: European Conference on Computer Vision. Springer International Publishing, 2016: 869-884.
• Linjie Yang, Ping Luo, Chen Change Loy, Xiaoou Tang. A Large-Scale Car Dataset for Fine-Grained Categorization and Verification, In Computer Vision and Pattern Recognition (CVPR), 2015.
• Vehicle Color Recognition using Convolutional Neural Network, Reza Fuad Rachmadi and I Ketut Eddy Purnama, https://arxiv.org/abs/1510.07391