Video Question Answering on VideoMME with subtitles
64.8Acc (Overall)HICom (Ours)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| HICom (Ours)LLM Size=7B, #Frames=128, #Tokens=2624, inference_frames_sampling_32=true2025.03 | 64.8 | 72.7 | 64.6 | 57.3 | |
| HICom (Ours)LLM Size=7B, #Frames=64, #Tokens=1328, inference_frames_sampling_32=true2025.03 | 63.3 | 71.8 | 63.7 | 54.5 | |
| LLaVA-VideoLLM Size=7B, #Frames=32, #Tokens=62722025.03 | 62.9 | 76.6 | 60.3 | 51.7 | |
| PAVE-7B (w/ video feature)FLOPs (TB)=98.63, Total/Trainable Params=8.2B/170.5M2025.03 | 62.9 | — | — | — | |
| LLaVA-OneVisionLLM Size=7B, #Frames=32, #Tokens=62722025.03 | 61.9 | 75.8 | 58.4 | 51.6 | |
| LLaVA-OV-7BFLOPs (TB)=98.53, Total/Trainable Params=8.2B/-2025.03 | 61.5 | — | — | — | |
| HICom (Ours)LLM Size=7B, #Frames=32, #Tokens=6802025.03 | 60.1 | 68.4 | 58.9 | 53.1 | |
| VITALLM Size=8x7B, #Frames=16, #Tokens=40962025.03 | 59.2 | 70.4 | 56.2 | 50.9 | |
| HICom (Ours)LLM Size=1.5B, #Frames=32, #Tokens=6802025.03 | 55.7 | 65.1 | 54.7 | 47.3 | |
| LongVALLM Size=7B, #Frames=128, #Tokens=184322025.03 | 54.3 | 61.6 | 53.6 | 47.6 | |
| PAVE-0.5B (w/ video feature)FLOPs (TB)=8.08, Total/Trainable Params=0.9B/41.4M2025.03 | 48.6 | — | — | — | |
| Chat-Univi-1.5LLM Size=7B, #Frames=64, #Tokens=4482025.03 | 45.9 | 51.2 | 44.6 | 41.8 | |
| VideoChat2-MistralLLM Size=7B, #Frames=16, #Tokens=15362025.03 | 43.8 | 52.8 | 39.4 | 39.2 | |
| LLaVA-OV-0.5BFLOPs (TB)=8.01, Total/Trainable Params=0.9B/-2025.03 | 43.5 | — | — | — | |
| Video-LLaVALLM Size=7B, #Frames=8, #Tokens=20482025.03 | 41.6 | 46.1 | 40.7 | 38.1 |