List your company
ResearchFormerly WizWand

Multi-choice Video Question Answering on EgoSchema (test)

72.2AccuracyGPT4-O
17.91232.00646.160.194Apr 9, 2024Jun 25, 2024Sep 10, 2024Nov 27, 2024Feb 12, 2025Apr 30, 2025Jul 17, 2025
Updated 7mo ago

Evaluation results

MethodLinks
GPT4-OVideo-Gen=false2024.1272.2
LongVU-7B2025.0767.6
VideoLLaMA2LLM size=72B, Video-Gen=false2024.1263.9
Gemini 1.5 ProVideo-Gen=false2024.1263.2
LLaVA-OneVisionLLM size=72B, Video-Gen=false2024.1262
LLaVA-NeXT-VideoLLM size=32B, Video-Gen=false2024.1260.9
LLaVA-OV-7B2025.0760.1
VILA-40B2025.0758
LLaVA-Video-7B2025.0757.3
Gemini 1.0 ProVideo-Gen=false2024.1255.7
GPT4-VVideo-Gen=false2024.1255.6
VideoLLaMA2LLM size=7B, Video-Gen=false2024.1251.7
MoReVQAVLM=PALI-3 (5B), LLM=PaLM-2, Video context=16 uniformly sampled frames2024.0451.7
InternVL2.5-8B-ITG-32Frames Sampled=322025.0751.6
InternVL2.5-8B2025.0751.5
VideoLLaMA2-7B2025.0750.5
JCEFVLM=PALI-3 (5B), LLM=PaLM-2, Video context=1 frame per second2024.0449.9
ViperGPT+VLM=PALI-3 (5B), LLM=PaLM-22024.0449.3
Divot-LLMLLM size=7B, Video-Gen=true2024.1246.5
LLaVA-NeXT-VideoLLM size=7B, Video-Gen=false2024.1243.9
VideoChat2LLM size=7B, Video-Gen=false2024.1242.2
LLM-onlyLLM=PaLM-22024.0441
LLaMA-VIDLLM size=7B, Video-Gen=false2024.1238.5
Video-LLaVALLM size=7B, Video-Gen=false2024.1238.4
Video-LaVITLLM size=7B, Video-Gen=true2024.1237.3
Random (for MC)2024.0420