TY - GEN
T1 - Semantic feature projection for continuous emotion analysis
AU - Lade, Prasanth
AU - McDaniel, Troy
AU - Panchanathan, Sethuraman
PY - 2014/11/3
Y1 - 2014/11/3
N2 - Affective computing researchers have recently been focusing on continuous emotion dimensions like arousal and valence. This dual coordinate affect space can explain many of the discrete emotions like sadness, anger, joy, etc. In the area of continuous emotion recognition, Principal Component Analysis (PCA) models are generally used to enhance the performance of various image and audio features by projecting them to a new space where the new features are less correlated. We instead, propose that quantizing and projecting the features to a latent topic space performs better than PCA. Specifically we extract these topic features using Latent Dirichlet Allocation (LDA) models. We show that topic models project the original features to a latent feature space that is more coherent and useful for continuous emotion recognition than PCA. Unlike PCA where no semantics can be attributed to the new features, topic features can have a visual and semantic interpretation which can be used in personalized HCI applications and Assistive technologies. Our hypothesis in this work has been validated using the AVEC 2012 continuous emotion challenge dataset.
AB - Affective computing researchers have recently been focusing on continuous emotion dimensions like arousal and valence. This dual coordinate affect space can explain many of the discrete emotions like sadness, anger, joy, etc. In the area of continuous emotion recognition, Principal Component Analysis (PCA) models are generally used to enhance the performance of various image and audio features by projecting them to a new space where the new features are less correlated. We instead, propose that quantizing and projecting the features to a latent topic space performs better than PCA. Specifically we extract these topic features using Latent Dirichlet Allocation (LDA) models. We show that topic models project the original features to a latent feature space that is more coherent and useful for continuous emotion recognition than PCA. Unlike PCA where no semantics can be attributed to the new features, topic features can have a visual and semantic interpretation which can be used in personalized HCI applications and Assistive technologies. Our hypothesis in this work has been validated using the AVEC 2012 continuous emotion challenge dataset.
KW - Feature comparison
KW - Topic models continuous affect recognition
UR - http://www.scopus.com/inward/record.url?scp=84913554573&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84913554573&partnerID=8YFLogxK
U2 - 10.1145/2647868.2655042
DO - 10.1145/2647868.2655042
M3 - Conference contribution
AN - SCOPUS:84913554573
T3 - MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia
SP - 881
EP - 884
BT - MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia
PB - Association for Computing Machinery
T2 - 2014 ACM Conference on Multimedia, MM 2014
Y2 - 3 November 2014 through 7 November 2014
ER -