###Joint sparse and low-rank representation for emotions.
Contact: [email protected]
Please download the DATA folder (processed Extended Cohn Kanade Database) from https://goo.gl/Wvfndv (Dropbox) or https://goo.gl/zt89pE (Google Drive) and place it inside icassp15 folder. In case you cannot access Dropbox or Google Drive, please download the CK+ dataset from http://www.consortium.ri.cmu.edu/ckagree/ and crop the faces using a detector such as Viola-Jones.
The program is developed on Windows. Except directory access issues such as '/' or '', 'ls' or 'dir' and so on, the program should be able to easily migarated to Linux. For details of the model and the algorithm, please refer to the paper available at https://arxiv.org/abs/1410.1606} or drop me a message. You are more than welcome to cite this work as
@inproceedings{xiang2015hierarchical,
title={Hierarchical Sparse and Collaborative Low-Rank Representation for Emotion Recognition},
author={Xiang, Xiang and Dao, Minh and Hager, Gregory D and Tran, Trac D},
booktitle={IEEE ICASSP},
pages={3811--3815},
year={2015},
}
@article{xiang2017linear, title={Linear Disentangled Representation Learning For Facial Actions}, author={Xiang, Xiang and Tran, Trac D}, journal={IEEE Transactions on Circuits and Systems for Video Technology}, year={2017}, publisher={IEEE} }
Xiang Xiang and Trac D. Tran: Linear Disentangled Representation Learning for Facial Actions. To appear at IEEE Trans. Circuits and System for Video Tech. (IEEE T-CSVT), Volume: PP, Issue: 99, 2018.
See also https://www.mathworks.com/matlabcentral/fileexchange/61323-eglxiang-icassp15-emotion for the comments from the Matlab developer community.