Domain Adaptation for Semantic Segmentation With Maximum Squares Loss

By Minghao Chen, Hongyang Xue, Deng Cai. IEEE International Conference on Computer Vision, 2019

320 citations recorded by the source corpus.

Deep neural networks for semantic segmentation always require a large number of samples with pixel-level labels, which becomes the major difficulty in their real-world applications. To reduce the labeling cost, unsupervised domain adaptation (UDA) approaches are proposed to transfer knowledge from labeled synthesized datasets to unlabeled real-world datasets. Recently, some semi-supervised learning methods have been applied to UDA and achieved state-of-the-art performance. One of the most popular approaches in semi-supervised learning is the entropy minimization method. However, when applying the entropy minimization to UDA for semantic segmentation, the gradient of the entropy is biased towards samples that are easy to transfer. To balance the gradient of well-classified target samples, we propose the maximum squares loss. Our maximum squares loss prevents the training process being dominated by easy-to-transfer samples in the target domain. Besides, we introduce the image-wise weighting ratio to alleviate the class imbalance in the unlabeled target domain. Both synthetic-to-real and cross-city adaptation experiments demonstrate the effectiveness of our proposed approach. The code is released at https://github. com/ZJULearning/MaxSquareLoss.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/af4082b90c5d71bb1f27b6ae2d197feca4d4778a

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Wen Li
2. Yahao Liu
3. Jinhong Deng
4. Lixin Duan
5. Jian Liang
6. Zhaoxiang Zhang
7. Yuxi Wang
8. Zhen Fang

In Computer Vision and Pattern Recognition

1. Luc Van Gool
2. Xiaogang Wang
3. Qi Tian
4. Dacheng Tao
5. Wen Li
6. Yi Yang
7. Trevor Darrell
8. Xiaoou Tang

Citations in the corpus

• Towards Discriminability and Diversity: Batch Nuclear-Norm Maximization Under Label Insufficient Situations (2020, background): “The entropy minimization is further modified into maximum squares loss in [6] to lower the influence of easy-to-transfer samples.”
• BAPA-Net: Boundary Adaptation and Prototype Alignment for Cross-domain Semantic Segmentation (2021, methodology): “A variety of unsupervised domain adaptation methods are proposed to maximally eliminate the domain discrepancy through adversarial feature learning [22, 6, 13], entropy minimization [52, 53, 5], selftraining [65, 66, 31, 11], etc.”
• Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation (2021, background): “Entropy minimization and its variants [9, 46, 57] encourage the network to make sharp predictions on the unlabeled target data, and the resulting network is prone to be over-confident on false predictions.”
• Source-Free Domain Adaptation for Semantic Segmentation (2021, background): “And LTAR is the self-supervision loss for the target domain based on pseudo-labels, such as entropy minimization [43], maximum square loss (MaxSquare) [6], etc.”
• PixMatch: Unsupervised Domain Adaptation via Pixelwise Consistency Training (2021, methodology): “Recently, a new line of work on UDA for semantic segmentation has emerged around self-training [63, 52, 7].”
• MetaCorrection: Domain-aware Meta Loss Correction for Unsupervised Domain Adaptation in Semantic Segmentation (2021, unknown): “Our MetaCorrection model, as a self-training based method, also outperforms other related pseudo label learning works [51, 4, 17, 23, 30, 46], demonstrating the effectiveness of the proposed method in alleviating the noise problem.”
• HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation (2022, unknown): “Several works also combine self-training and adversarial training [34, 37, 64, 86], minimize the entropy [7, 62, 63], refine boundaries [39], use a curriculum [12, 13, 82], or learn auxiliary tasks [8, 29, 63, 66].”
• Where and How to Transfer: Knowledge Aggregation-Induced Transferability Perception for Unsupervised Domain Adaptation (2021, background): “Some efficient objectives for knowledge adaptation [6] (e.g., sliced Wasserstein discrepancy [24], maximum squares loss [4], adversarial entropy minimization [43] and scale-invariant formulation [41]) are designed to narrow the domain gap.”
• A Comprehensive Survey on Test-Time Adaptation Under Distribution Shifts (2023, unknown): “, inputlevel translation [28], [29], feature-level alignment [30], [31], [32]), output-level regularization [33], [34], [35], and prior estimation [17], [36], [37].”
• Give Me Your Trained Model: Domain Adaptive Semantic Segmentation without Source Data (2021, methodology): “To remedy this issue, various domain alignment strategies are proposed in [8, 2, 9–11].”