Constructing Holistic Spatio-Temporal Scene Graph for Video Semantic Role Labeling.
Yu ZhaoHao FeiYixin CaoBobo LiMeishan ZhangJianguo WeiMin ZhangTat-Seng ChuaPublished in: ACM Multimedia (2023)
Keyphrases
- semantic role labeling
- spatio temporal
- video sequences
- moving objects
- image sequences
- space time
- dynamic textures
- syntactic parsing
- dependency parsing
- relation extraction
- semantic roles
- video data
- predicate argument
- tree kernels
- video frames
- structured data
- named entity recognition
- word sense disambiguation
- information extraction
- key frames
- syntactic dependencies
- automatic extraction
- low level
- search engine