LocVTP: Video-Text Pre-training for Temporal Localization.
Meng CaoTianyu YangJunwu WengCan ZhangJue WangYuexian ZouPublished in: CoRR (2022)
Keyphrases
- temporal information
- spatial and temporal
- activity detection
- temporal consistency
- video sequences
- temporal correlation
- temporal coherence
- natural language descriptions
- spatial temporal
- space time
- temporal expressions
- real time
- video streams
- video search
- temporal domain
- video data
- news video
- video frames
- spatio temporal
- temporal structure
- temporal resolution
- information retrieval
- training process
- text mining
- training set
- training examples
- text detection
- temporal data
- multimedia
- temporal analysis
- text documents
- video content
- video database
- text retrieval
- video segments
- multimedia documents
- multimedia search
- temporal constraints
- combining information from multiple
- lecture videos
- object based video
- temporal dimension
- temporal relationships
- training corpus
- textual descriptions
- video shots
- video analysis
- video clips
- semantic information
- text classification