eglxiang/icassp15_e​motion

This reoisitory contains the program assiciated with the paper with ID SPTM-P8.5 at IEEE ICASSP 2015

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. 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 International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
pages={3811--3815},
year={2015},
organization={IEEE}
}