Multimodal Sentiment Analysis on CH-SIMSv2 ID (test)
53.6Accuracy (5-class)Qwen2.5Omni
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Qwen2.5OmniTraining dataset=SIMS2026.03 | 53.6 | 75.3 | 85.4 | 85.2 | 0.263 | 0.833 | |
| VideoLLaMA2Training dataset=SIMS2026.03 | 43.3 | 76.4 | 80.9 | 80.9 | 0.494 | 0.726 | |
| HumanOmniTraining dataset=SIMS2026.03 | 43.3 | 73.1 | 81.5 | 81.3 | 0.381 | 0.752 | |
| MSA-ThinkerTraining dataset=SIMS2026.03 | 42.6 | 80 | 84.9 | 85 | 0.407 | 0.796 | |
| PandaGPTTraining dataset=SIMS2026.03 | 40 | 57.6 | 70.4 | 67.9 | 0.403 | 0.55 | |
| OlaTraining dataset=SIMS2026.03 | 39.8 | 70.7 | 77.6 | 76.6 | 0.46 | 0.645 | |
| Emotion-LLaMATraining dataset=SIMS2026.03 | 39.4 | 57.4 | 72.3 | 70.5 | 0.402 | 0.61 | |
| MiniCPM-oTraining dataset=SIMS2026.03 | 39.1 | 68.5 | 76.6 | 75 | 0.42 | 0.7 |