TimeChat-Online: 80% Visual Tokens are Naturally Redundant in Streaming Videos
Open paper
Benchmark
Model
Rank
Results
video-question-answering-on-ovo-bench
TimeChat-Online-7B
–
Real-Time Visual Perception: 61.9
Backward Tracing: 41.7
…
video-question-answering-on-streamingbench
TimeChat-Online-7B
–
Real-Time Visual Understanding: 75.28
Rank counts only results with a code link.