Text-to-audio-video generation on JavisBench mini (test)
317.5FVDJavisGPT
Evaluation Results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| JavisGPTModel Architecture Type=MLLM2025.12 | 317.5 | 1.8 | 7.6 | 0.145 | 0.18 | 0.324 | 0.308 | 0.202 | 0.797 | 0.185 | 0.157 | |
| JavisDiTModel Architecture Type=DiT2025.12 | 327.8 | 1.9 | 7.6 | 0.141 | 0.184 | 0.322 | 0.314 | 0.203 | 0.799 | 0.181 | 0.153 | |
| NExT-GPTModel Architecture Type=MLLM2025.12 | 1,463.2 | 6.1 | 7 | 0.123 | 0.153 | 0.207 | 0.306 | 0.067 | 0.732 | 0.061 | 0.038 | |
| UnifiedIO-2Model Architecture Type=MLLM2025.12 | 1,597.4 | 6.4 | 6.7 | 0.137 | 0.138 | 0.313 | 0.294 | 0.106 | 0.747 | 0.094 | 0.053 | |
| MM-DiffModel Architecture Type=DiT2025.12 | 2,311.9 | 12.2 | 27.5 | 0.08 | 0.032 | 0.173 | 0.048 | 0.119 | 0.783 | 0.109 | 0.07 |