ModelScope

Reported on 5 benchmarks across 1 task · 1 paper · 5 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing5 results

Text-to-Video GenerationonEvalCrafter Text-to-Video (ECTV) Dataset
Motion Quality· uses extra data· 2023-06-03
53.09
best: 63.98 (VideoCrafter2)
SOTA
VideoComposer: Compositional Video Synthesis with Motion Controllability arXiv:2306.02018
Text-to-Video GenerationonEvalCrafter Text-to-Video (ECTV) Dataset
Temporal Consistency· uses extra data· 2023-06-03
54.46
best: 61.46 (VideoCrafter2)
SOTA
VideoComposer: Compositional Video Synthesis with Motion Controllability arXiv:2306.02018
Text-to-Video GenerationonEvalCrafter Text-to-Video (ECTV) Dataset
Text-to-Video Alignment· uses extra data· 2023-06-03
57.8
best: 68.49 (Lavie)
SOTA
VideoComposer: Compositional Video Synthesis with Motion Controllability arXiv:2306.02018
Text-to-Video GenerationonEvalCrafter Text-to-Video (ECTV) Dataset
Total Score· uses extra data· 2023-06-03
218
best: 243 (VideoCrafter2)
SOTA
VideoComposer: Compositional Video Synthesis with Motion Controllability arXiv:2306.02018
Text-to-Video GenerationonEvalCrafter Text-to-Video (ECTV) Dataset
Visual Quality· uses extra data· 2023-06-03
52.47
best: 54.82 (VideoCrafter2)
SOTA
VideoComposer: Compositional Video Synthesis with Motion Controllability arXiv:2306.02018