List your company
ResearchFormerly WizWand

Egocentric Action Planning on EgoPlan

38AccuracyGPT-4v
27.0829.91532.7535.585Mar 5, 2025May 8, 2025Jul 12, 2025Sep 15, 2025Nov 18, 2025Jan 22, 2026Mar 28, 2026
Updated 7mo ago

Evaluation results

MethodLinks
GPT-4v#Frames=322025.0338
TGPOBackbone=Qwen2.5-VL 3B2026.0336.8
EgoVLM GRPO (3B)*Model Scale=3B, Training Protocol=GRPO, Model Category=Open-source MLLMs2026.0336.5
Qwen2-VL#Param=7B, #Frames=322025.0334.3
LLaVA-Videos#Param=7B, #Frames=322025.0333.6
LLaVA-VideoModel Scale=7B, Model Category=Open-source MLLMs2026.0333.6
EgoGPT#Param=7B, #Frames=32, training_dataset=EgoIT+EgoLifeD12025.0333.4
Oryx#Param=7B, #Frames=322025.0333.2
Qwen2.5-VLModel Scale=7B, Model Category=Open-source MLLMs2026.0333
EgoVLM Dr. GRPOModel Scale=3B, Training Protocol=Dr. GRPO, Model Category=Open-source MLLMs2026.0333
Qwen2.5-VLModel Scale=3B, Model Category=Open-source MLLMs2026.0332.9
GPT-4o#Frames=322025.0332.8
Gemini 1.5-ProModel Category=Proprietary2026.0332.8
GPT-4oModel Category=Proprietary2026.0332.8
EgoGPT#Param=7B, #Frames=32, training_dataset=EgoIT2025.0332.4
EgoVLM SFT (3B)*Model Scale=3B, Training Protocol=SFT, Model Category=Open-source MLLMs2026.0332.1
Gemini-1.5-Pro#Frames=322025.0331.3
LLaVA-OV#Param=7B, #Frames=322025.0330.7
LongVA#Param=7B, #Frames=322025.0329.9
IXC-2.5#Param=7B, #Frames=322025.0329.4
LLaVA-Next-Video#Param=7B, #Frames=322025.0329
InternVideo2#Param=8B, #Frames=322025.0327.5