T2V-CompBench: A Comprehensive Benchmark for Compositional Text-to-video Generation.
Kaiyue SunKaiyi HuangXian LiuYue WuZihan XuZhenguo LiXihui LiuPublished in: CoRR (2024)
Keyphrases
- text generation
- video data
- free text
- text retrieval
- natural language descriptions
- video streams
- video sequences
- text detection
- space time
- video segments
- real time video
- real time
- video content
- video frames
- multimedia
- spatial and temporal
- audio content
- news video
- information retrieval
- video search
- multimedia search
- video images
- textual data
- online video
- video collections
- natural language generation
- closed captions
- multimedia documents
- video shots
- video clips
- web documents
- real world
- video analysis
- video scene
- key frames
- temporal information
- text documents
- semantic labels
- motion estimation
- spatio temporal
- natural language
- keywords