Multimodal Sentiment Analysis on CMU-MOSEI v1 (test)
87.3Accuracy (2-Class)UMQ
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| UMQNoise Type=Gaussian Noise, NR=0.12026.03 | 87.3 | 0.531 | |
| UMQ (OOD)Noise Type=Other Noises (Laplace and Random Erasing), NR=0.12026.03 | 87 | 0.518 | |
| UMQNoise Type=Gaussian Noise, NR=Avg2026.03 | 86.5 | 0.551 | |
| UMQ (OOD)Noise Type=Other Noises (Laplace and Random Erasing), NR=Avg2026.03 | 86.3 | 0.536 | |
| C-MIBNoise Type=Gaussian Noise, NR=0.12026.03 | 86.1 | 0.545 | |
| NIATNoise Type=Gaussian Noise, NR=0.12026.03 | 80.3 | 0.554 |