Video Question Answering on EgoSchema Sub.
79.4AccuracyTASKER
Evaluation Results
| Method | Links | |
|---|---|---|
| TASKER(M)LLM=Qwen3-VL2026.06 | 79.4 | |
| VideoTree(M)LLM=Qwen3-VL2026.06 | 77.2 | |
| VideoAgent(M)LLM=Qwen3-VL2026.06 | 75.8 | |
| TASKER(M)LLM=GPT-4o2026.06 | 68.6 | |
| LVNet(M)LLM=GPT-4o2026.06 | 68.2 | |
| TASKER(M)LLM=GPT-42026.06 | 68 | |
| LangRepo(M)LLM=Mixtral-8x7B2026.06 | 66.2 | |
| VideoTree(M)LLM=GPT-42026.06 | 66.2 | |
| LifelongMemory(M)LLM=GPT-42026.06 | 64.1 | |
| VideoAgent(M)LLM=GPT-42026.06 | 62.8 | |
| LLoVi(M)LLM=GPT-42026.06 | 61.2 | |
| MVU(M)LLM=Mistral-13B2026.06 | 60.3 | |
| VideoAgent(M)LLM=GPT-42026.06 | 60.2 | |
| IG-VLM(M)LLM=GPT-4V2026.06 | 59.8 | |
| ProViQ(M)LLM=GPT-3.52026.06 | 57.1 | |
| Vamos(M)LLM=GPT-42026.06 | 51.2 |