Multimodal Sentiment Analysis on CMU-MOSEI OOD (test)
49.3Accuracy (7-class)Qwen2.5Omni
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Qwen2.5OmniTraining Dataset=SIMS, Distribution=OOD, Model Size=7B2026.03 | 49.3 | 84.1 | 83.4 | 0.637 | 0.688 | |
| OlaTraining Dataset=SIMS, Distribution=OOD, Model Size=7B2026.03 | 47.4 | 83.2 | 83.2 | 0.666 | 0.672 | |
| MiniCPM-oTraining Dataset=SIMS, Distribution=OOD, Model Size=7B2026.03 | 47.1 | 85.3 | 84.9 | 0.641 | 0.688 | |
| MSA-Thinker(SIMS)Training Dataset=SIMS, Distribution=OOD, Model Size=7B2026.03 | 46.3 | 86.6 | 86.3 | 0.635 | 0.731 | |
| PandaGPTTraining Dataset=SIMS, Distribution=OOD, Model Size=7B2026.03 | 44.6 | 85.5 | 85.3 | 0.723 | 0.61 | |
| HumanOmniTraining Dataset=SIMS, Distribution=OOD, Model Size=7B2026.03 | 44 | 84.8 | 84.6 | 0.682 | 0.63 | |
| Emotion-LLaMATraining Dataset=SIMS, Distribution=OOD, Model Size=7B2026.03 | 38.5 | 80.7 | 80.5 | 1.224 | 0.025 | |
| VideoLLaMA2Training Dataset=SIMS, Distribution=OOD, Model Size=7B2026.03 | 37.7 | 74.4 | 72 | 0.835 | 0.448 |