TY - GEN
T1 - Complexity-aware assignment of latent values in discriminative models for accurate gesture recognition
AU - Ribeiro, Manoel Horta
AU - Teixeira, Bruno
AU - Fernandes, Antonio Otavio
AU - Meira, Wagner
AU - Nascimento, Erickson R.
N1 - Publisher Copyright:
© 2016 IEEE.
PY - 2017/1/10
Y1 - 2017/1/10
N2 - Many of the state-of-the-art algorithms for gesture recognition are based on Conditional Random Fields (CRFs). Successful approaches, such as the Latent-Dynamic CRFs, extend the CRF by incorporating latent variables, whose values are mapped to the values of the labels. In this paper we propose a novel methodology to set the latent values according to the gesture complexity. We use an heuristic that iterates through the samples associated with each label value, estimating their complexity. We then use it to assign the latent values to the label values. We evaluate our method on the task of recognizing human gestures from video streams. The experiments were performed in binary datasets, generated by grouping different labels. Our results demonstrate that our approach outperforms the arbitrary one in many cases, increasing the accuracy by up to 10%.
AB - Many of the state-of-the-art algorithms for gesture recognition are based on Conditional Random Fields (CRFs). Successful approaches, such as the Latent-Dynamic CRFs, extend the CRF by incorporating latent variables, whose values are mapped to the values of the labels. In this paper we propose a novel methodology to set the latent values according to the gesture complexity. We use an heuristic that iterates through the samples associated with each label value, estimating their complexity. We then use it to assign the latent values to the label values. We evaluate our method on the task of recognizing human gestures from video streams. The experiments were performed in binary datasets, generated by grouping different labels. Our results demonstrate that our approach outperforms the arbitrary one in many cases, increasing the accuracy by up to 10%.
KW - activity recognition
KW - conditional random fields
KW - discriminative models
KW - gesture recognition
UR - http://www.scopus.com/inward/record.url?scp=85013805800&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85013805800&partnerID=8YFLogxK
U2 - 10.1109/SIBGRAPI.2016.059
DO - 10.1109/SIBGRAPI.2016.059
M3 - Conference contribution
AN - SCOPUS:85013805800
T3 - Proceedings - 2016 29th SIBGRAPI Conference on Graphics, Patterns and Images, SIBGRAPI 2016
SP - 378
EP - 385
BT - Proceedings - 2016 29th SIBGRAPI Conference on Graphics, Patterns and Images, SIBGRAPI 2016
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 29th SIBGRAPI Conference on Graphics, Patterns and Images, SIBGRAPI 2016
Y2 - 4 October 2016 through 7 October 2016
ER -