List your company

Open-ended Video Question Answering on MovieChat-1K Global mode

93.1AccuracyDrVideo
45.7858.06570.3582.635Jun 18, 2024

Evaluation results

MethodLinks
DrVideoArchitecture=Based on Open-source Captioners and Proprietary LLMs2024.0693.14.41
Sullam Jeoung, et alArchitecture=Based on Open-source Captioners and Proprietary LLMs2024.0684.8—
MovieChat+Architecture=Open-Source MLLM2024.0671.23.51
MovieChatArchitecture=Open-Source MLLM2024.0668.33.15
VideoAgentArchitecture=Based on Open-source Captioners and Proprietary LLMs, re-implemented=true2024.0665.43.12
MM-VIDArchitecture=Based on Open-source Captioners and Proprietary LLMs2024.0658.62.86
LLoViArchitecture=Based on Open-source Captioners and Proprietary LLMs2024.0658.32.87
Video ChatArchitecture=Open-Source MLLM2024.0657.83.08
Video LLaMAArchitecture=Open-Source MLLM2024.0651.72.72
Video-ChatGPTArchitecture=Open-Source MLLM2024.0647.62.89