Genotype Prediction on FMR1
68.67AccuracyRaw pose + SVM
Evaluation Results
| Method | Links | |
|---|---|---|
| Raw pose + SVMRepresentation=Joint coordinates, Evaluation Protocol=Hand-crafted Features + Linear Probe2026.05 | 68.67 | |
| PCA + SVMRepresentation=PCA components, Evaluation Protocol=Hand-crafted Features + Linear Probe2026.05 | 67.79 | |
| GEESERepresentation=Fine-tuned embedding, Evaluation Protocol=End-to-End Fine-tuned (behavior training → genotype fine-tuning)2026.05 | 67.7 | |
| s-DANNCERepresentation=Wavelet + PCA (421-dim), Evaluation Protocol=Hand-crafted Features + Linear Probe2026.05 | 66.45 | |
| CEBRARepresentation=Contrastive embedding, Evaluation Protocol=Learned Representations + Linear Probe2026.05 | 53.43 | |
| Random GuessEvaluation Protocol=Reference2026.05 | 50 | |
| MOMENT (frozen)Representation=Pretrained embedding, Evaluation Protocol=Learned Representations + Linear Probe2026.05 | 44.2 |