Adaptive Semi-Supervised Graph Contrastive Representation Learning for Robust Emotion Recognition Using Electroencephalogram
Published
Research approach
Updated sample graphs with learned representations and used label-guided augmentation to reduce noisy connections.
View full-size figure
View full-size figure
View full-size figure
Results
SEED-IV Case 1: 75.75 ± 8.95% for Static GCN and 87.31 ± 3.84% for ASGCRL, a difference of 11.56 percentage points under the same protocol.
| SEED-IV Case 1 | Accuracy |
|---|---|
| Static GCN | 75.75 ± 8.95% |
| ASGCRL | 87.31 ± 3.84% |
| Matched-protocol difference | +11.56%p |
InterpretationSEED-IV Case 1, 2.4% labels, subject-wise transductive evaluation. Embedding example: subject 15.
All publications