Emotion recognition in the wild from videos using images
Authors:
Open Access
- 31 October 2016
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 433-436
Abstract
This paper presents the implementation details of the proposed solution to the Emotion Recognition in the Wild 2016 Challenge, in the category of video-based emotion recognition. The proposed approach takes the video stream from the audio-video trimmed clips provided by the challenge as input and produces the emotion label corresponding to this video sequence. This output is encoded as one out of seven classes: the six basic emotions (Anger, Disgust, Fear, Happiness, Sad, Surprise) and Neutral. Overall, the system consists of several pipelined modules: face detection, image pre-processing, deep feature extraction, feature encoding and, finally, an SVM classification. This system achieves 59.42% validation accuracy, surpassing the competition baseline of 38.81%. With regard to test data, our system achieves 56.66% recognition rate, also improving the competition baseline of 40.47%.Keywords
This publication has 17 references indexed in Scilit:
- Training deep networks for facial expression recognition with crowd-sourced label distribution2016
- EmotiW 2016: video and group-level emotion recognition challenges2016
- Recurrent Neural Networks for Emotion Recognition in Video2015
- Automatic measurement of ad preferences from facial responses gathered over the InternetImage and Vision Computing, 2014
- Large-Scale Video Classification with Convolutional Neural Networks2014
- Collecting Large, Richly Annotated Facial-Expression Databases from MoviesIEEE MultiMedia, 2012
- ImageNet: A large-scale hierarchical image database2009
- Real-Time Nonintrusive Monitoring and Prediction of Driver FatigueIEEE Transactions on Vehicular Technology, 2004
- Automatic facial expression analysis: a surveyPattern Recognition, 2003
- Facial Action Coding System1978