Semantic Segmentation on Cityscapes (R1, R0, Rt, Rr, 1/D1)
2.113R1Independent
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| IndependentCodec=Independent, eta=0.0012026.01 | 2.113 | 0 | 4.409 | 4.409 | 0.666 | |
| IndependentCodec=Independent, eta=0.012026.01 | 1.808 | 0 | 3.7 | 3.7 | 0.668 | |
| Learnable Gray-Wyner NetworkCodec=Proposed, eta=0.0012026.01 | 1.384 | 0.33 | 3.196 | 3.526 | 0.666 | |
| Learnable Gray-Wyner NetworkCodec=Proposed, eta=0.012026.01 | 1.329 | 0.243 | 2.987 | 3.23 | 0.669 | |
| IndependentCodec=Independent, eta=0.052026.01 | 1.227 | 0 | 2.564 | 2.564 | 0.665 | |
| Learnable Gray-Wyner NetworkCodec=Proposed, eta=0.052026.01 | 1.217 | 0.329 | 2.831 | 3.16 | 0.669 | |
| Learnable Gray-Wyner NetworkCodec=Proposed, eta=0.12026.01 | 1.11 | 0.414 | 2.675 | 3.089 | 0.669 | |
| IndependentCodec=Independent, eta=0.12026.01 | 1.104 | 0 | 2.295 | 2.295 | 0.663 | |
| IndependentCodec=Independent, eta=12026.01 | 0.788 | 0 | 1.634 | 1.634 | 0.66 | |
| Learnable Gray-Wyner NetworkCodec=Proposed, eta=12026.01 | 0.568 | 0.409 | 1.545 | 1.954 | 0.666 | |
| Learnable Gray-Wyner NetworkCodec=Proposed, eta=52026.01 | 0.315 | 0.319 | 0.945 | 1.264 | 0.662 | |
| JointCodec=Joint, eta=0.0012026.01 | 0 | 2.444 | 2.444 | 4.888 | 0.67 | |
| JointCodec=Joint, eta=0.012026.01 | 0 | 2.069 | 2.069 | 4.138 | 0.669 | |
| JointCodec=Joint, eta=0.052026.01 | 0 | 1.611 | 1.611 | 3.222 | 0.67 | |
| JointCodec=Joint, eta=0.12026.01 | 0 | 1.527 | 1.527 | 3.054 | 0.669 | |
| JointCodec=Joint, eta=12026.01 | 0 | 1.074 | 1.074 | 2.148 | 0.663 |