Unsupervised Scene Adaptation with Memory Regularization in vivo
About
We consider the unsupervised scene adaptation problem of learning from both labeled source data and unlabeled target data. Existing methods focus on minoring the inter-domain gap between the source and target domains. However, the intra-domain knowledge and inherent uncertainty learned by the network are under-explored. In this paper, we propose an orthogonal method, called memory regularization in vivo to exploit the intra-domain knowledge and regularize the model training. Specifically, we refer to the segmentation model itself as the memory module, and minor the discrepancy of the two classifiers, i.e., the primary classifier and the auxiliary classifier, to reduce the prediction inconsistency. Without extra parameters, the proposed method is complementary to the most existing domain adaptation methods and could generally improve the performance of existing methods. Albeit simple, we verify the effectiveness of memory regularization on two synthetic-to-real benchmarks: GTA5 -> Cityscapes and SYNTHIA -> Cityscapes, yielding +11.1% and +11.3% mIoU improvement over the baseline model, respectively. Besides, a similar +12.0% mIoU improvement is observed on the cross-city benchmark: Cityscapes -> Oxford RobotCar.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Semantic segmentation | GTA5 → Cityscapes (val) | mIoU48.3 | 533 | |
| Semantic segmentation | SYNTHIA to Cityscapes (val) | Rider IoU25.1 | 435 | |
| Semantic segmentation | GTA5 to Cityscapes (test) | mIoU48.3 | 151 | |
| Semantic segmentation | SYNTHIA to Cityscapes | Road IoU83.1 | 150 | |
| Semantic segmentation | Synthia to Cityscapes (test) | Road IoU82 | 138 | |
| Semantic segmentation | Cityscapes (val) | mIoU50.3 | 133 | |
| Semantic segmentation | Cityscapes adaptation from Synthia 1.0 (val) | Person IoU61.3 | 114 | |
| Semantic segmentation | GTA5 to Cityscapes 1.0 (val) | Road IoU90.5 | 98 | |
| Semantic segmentation | GTA to Cityscapes | Road IoU89.1 | 72 | |
| Semantic segmentation | GTA5 to Cityscapes | mIoU48.3 | 58 |