Open Ended Question Answering on MSVD
73.92AccuracyMiniGPT4-Video
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MiniGPT4-VideoBackbone=Mistral-7B, Using Subtitles=false, Zero-shot=true2024.04 | 73.92 | 4.06 | |
| MiniGPT4-VideoBackbone=Llama 2-7B, Using Subtitles=false, Zero-shot=true2024.04 | 72.93 | 3.84 | |
| VideoLLaMA 2Model Size=7B, # Frames=82024.06 | 71.7 | 3.9 | |
| VideoLLaMA 2Model Size=72B, # Frames=82024.06 | 71 | 3.8 | |
| VideoLLaMA 2Model Size=7B, # Frames=162024.06 | 70.9 | 3.8 | |
| Video-LLaVAModel Size=7B, # Frames=82024.06 | 70.7 | 3.9 | |
| VideoLLaMA 2.1Model Size=7B, # Frames=162024.06 | 70.6 | 3.8 | |
| VideoLLaMA 2Model Size=8x7B, # Frames=82024.06 | 70.5 | 3.8 | |
| VideoChat2Model Size=7B, # Frames=162024.06 | 70 | 3.9 | |
| LLaMA-VID-7BUsing Subtitles=false, Zero-shot=true2024.04 | 69.7 | 3.7 | |
| LLaMA-VIDModel Size=7B, # Frames=1 fps2024.06 | 69.7 | 3.7 | |
| LLaVA-NeXT-VideoModel Size=7B, # Frames=322024.06 | 67.8 | 3.5 | |
| BT-Adapter-7BUsing Subtitles=false, Zero-shot=true2024.04 | 67.7 | 3.7 | |
| Chat-UniViModel Size=7B, # Frames=82024.06 | 65 | 3.6 | |
| Video-ChatGPTUsing Subtitles=false, Zero-shot=true2024.04 | 64.9 | 3.3 | |
| Video-ChatGPTModel Size=7B, # Frames=82024.06 | 64.9 | 3.3 | |
| Video ChatUsing Subtitles=false, Zero-shot=true2024.04 | 56.3 | 2.8 | |
| VideoChatModel Size=7B, # Frames=82024.06 | 56.3 | 2.8 | |
| LLaMA AdapterUsing Subtitles=false, Zero-shot=true2024.04 | 54.9 | 3.1 | |
| Video LLaMAUsing Subtitles=false, Zero-shot=true2024.04 | 51.6 | 2.5 | |
| VideoLLaMAModel Size=7B, # Frames=82024.06 | 51.6 | 2.5 | |
| FrozenBiLMUsing Subtitles=false, Zero-shot=true2024.04 | 32.2 | — |