Enhancing Transformer Backbone for Egocentric Video Action Segmentation.
Sakib RezaBalaji SundareshanMohsen MoghaddamOctavia I. CampsPublished in: CoRR (2023)
Keyphrases
- video segmentation
- human actions
- video data
- video sequences
- video streams
- video summarization
- medical images
- segmentation algorithm
- video content
- level set
- temporal segmentation
- action classification
- video frames
- text detection
- action recognition
- video analysis
- multiscale
- image segmentation
- segmentation method
- story segmentation
- object segmentation and tracking
- motion segmentation
- region growing
- space time
- shape prior
- dynamic scenes
- activity recognition
- foreground background segmentation
- segmentation accuracy
- video objects
- image analysis
- real time
- layered representation
- action detection
- fuzzy logic
- natural images
- video retrieval
- video scene
- edge detection
- news video
- visual saliency
- computer vision
- surveillance videos
- recognizing human actions
- video clips