CO-LDA: A Semi-supervised Approach to Audio-Visual Person Recognition.
Xuran ZhaoNicholas W. D. EvansJean-Luc DugelayPublished in: ICME (2012)
Keyphrases
- audio visual
- semi supervised
- multi modal
- feature extraction
- visual information
- multi stream
- object recognition
- multimedia
- face recognition
- pattern recognition
- topic models
- audio features
- machine learning
- labeled data
- action recognition
- activity recognition
- visual data
- active learning
- pairwise
- image representation
- human body
- training data
- spatio temporal