List your company

Video Question Answering on Vript RR

0.921Scene Accuracy (M)GPT-4O
0.31260.470550.62850.78645Jun 10, 2024

Evaluation results

MethodLinks
GPT-4O2024.060.9210.7750.7240.546
GPT-4OSampling rate=1 fps/100 fr2024.060.9140.7820.7850.66
Gemini-1.5-Pro2024.060.8510.6850.5990.475
GPT-4V2024.060.8090.750.7170.61
Claude 3.5-Sonnet2024.060.8090.5920.5460.428
VILA-1.5Parameters=8B2024.060.750.4870.5530.323
Claude 3-Opus2024.060.6380.60520.6050.434
PLLAVAParameters=7B2024.060.6250.4610.5530.362
Claude 3-Sonnet2024.060.6050.5390.5660.421
VideoChat22024.060.520.3220.4210.224
ST-LLM2024.060.4340.3490.3360.263
Video-LLaMA2024.060.3820.1970.2830.145
VideoChatGPT2024.060.3420.2890.2960.178
VideoChat2024.060.3360.230.2240.151