Unsupervised Scene Adaptation with Memory Regularization in vivo

24 Dec 2019  ยท  Zhedong Zheng, Yi Yang ยท

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.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Unsupervised Domain Adaptation Cityscapes-to-OxfordCar MRNet mIoU 73.9 # 3
Domain Adaptation GTA5+Synscapes to Cityscapes MRNet mIoU 47.6 # 5
Domain Adaptation GTA5 to Cityscapes MRNet mIoU 48.3 # 25
Domain Adaptation GTAV+Synscapes to Cityscapes MRNet mIoU 47.6 # 6
Synthetic-to-Real Translation GTAV-to-Cityscapes Labels MRNet mIoU 48.3 # 51
Unsupervised Domain Adaptation GTAV-to-Cityscapes Labels MRNet mIoU 45.5 # 20
Unsupervised Domain Adaptation SYNTHIA-to-Cityscapes MRNet mIoU (13 classes) 50.2 # 18
mIoU 43.2 # 13
Synthetic-to-Real Translation SYNTHIA-to-Cityscapes MRNet(ResNet-101) MIoU (13 classes) 53.8 # 27
Domain Adaptation SYNTHIA-to-Cityscapes Labels MRNet mIoU 46.5 # 1

Methods