Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain Adaptive Semantic Segmentation

Abstract

This paper focuses on the unsupervised domain adaptation of transferring the knowledge from the source domain to the target domain in the context of semantic segmentation. Existing approaches usually regard the pseudo label as the ground truth to fully exploit the unlabeled target-domain data. Yet the pseudo labels of the target-domain data are usually predicted by the model trained on the source domain. Thus, the generated labels inevitably contain the incorrect prediction due to the discrepancy between the training domain and the test domain, which could be transferred to the final adapted model and largely compromises the training process. To overcome the problem, this paper proposes to explicitly estimate the prediction uncertainty during training to rectify the pseudo label learning for unsupervised semantic segmentation adaptation. Given the input image, the model outputs the semantic segmentation prediction as well as the uncertainty of the prediction. Specifically, we model the uncertainty via the prediction variance and involve the uncertainty into the optimization objective. To verify the effectiveness of the proposed method, we evaluate the proposed method on two prevalent synthetic-to-real semantic segmentation benchmarks, i.e., GTA5 -> Cityscapes and SYNTHIA -> Cityscapes, as well as one cross-city benchmark, i.e., Cityscapes -> Oxford RobotCar. We demonstrate through extensive experiments that the proposed approach (1) dynamically sets different confidence thresholds according to the prediction variance, (2) rectifies the learning from noisy pseudo labels, and (3) achieves significant improvements over the conventional pseudo label learning and yields competitive performance on all three benchmarks.

13 pages, 6 figures, 10 tables, accepted by IJCV

References in corpus (7)

• Training Deep Neural Networks on Noisy Labels with Bootstrapping
• Playing for Data: Ground Truth from Computer Games
• Pixel-Level Cycle Association: A New Perspective for Domain Adaptive Semantic Segmentation
• Adversarial Style Mining for One-Shot Unsupervised Domain Adaptation
• Transferrable Prototypical Networks for Unsupervised Domain Adaptation
• Significance-aware Information Bottleneck for Domain Adaptive Semantic Segmentation
• ACE: Adapting to Changing Environments for Semantic Segmentation

Cited by in corpus (7)

• Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation
• Efficient Semi-Supervised Gross Target Volume of Nasopharyngeal Carcinoma Segmentation via Uncertainty Rectified Pyramid Consistency
• Uncertainty-Guided Mixup for Semi-Supervised Domain Adaptation without Source Data
• Unsupervised Semantic Aggregation and Deformable Template Matching for Semi-Supervised Learning
• Exploiting Sample Uncertainty for Domain Adaptive Person Re-Identification
• Self-Paced Uncertainty Estimation for One-shot Person Re-Identification
• Dynamic Domain Adaptation for Efficient Inference