EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action Recognition.
Evangelos KazakosArsha NagraniAndrew ZissermanDima DamenPublished in: CoRR (2019)
Keyphrases
- action recognition
- audio visual
- activity recognition
- video summarization
- multimodal fusion
- person authentication
- multi modal
- human actions
- visual data
- motion history images
- visual information
- computer vision
- action classification
- multimedia
- bag of words
- temporal information
- multi stream
- spatial and temporal
- recognizing human actions
- spatio temporal
- space time
- human activities
- multiscale
- human motion
- temporal relations
- atomic actions
- object recognition
- three dimensional