TY - JOUR
T1 - Robust unsupervised domain adaptation by retaining confident entropy via edge concatenation
AU - Hong, Hye Seong
AU - Kumar, Abhishek
AU - Lee, Dong Gyu
N1 - Publisher Copyright:
© 2023 Elsevier Ltd
PY - 2024/3/15
Y1 - 2024/3/15
N2 - The generalization capability of unsupervised domain adaptation can mitigate the need for extensive pixel-level annotations to train semantic segmentation networks by training models on synthetic data as a source with computer-generated annotations. Entropy-based adversarial networks are proposed to improve source domain prediction; however, they disregard significant external information, such as edges, which have the potential to identify and distinguish various objects within an image accurately. To address this issue, we introduce a novel approach to domain adaptation, leveraging the synergy of internal and external information within entropy-based adversarial networks. In this approach, we enrich the discriminator network with edge-predicted probability values within this innovative framework to enhance the clarity of class boundaries. Furthermore, we devised a probability-sharing network that integrates diverse information for more effective segmentation. Incorporating object edges addresses a pivotal aspect of unsupervised domain adaptation that has frequently been neglected in the past—the precise delineation of object boundaries. Conventional unsupervised domain adaptation methods usually center around aligning feature distributions and may not explicitly model object boundaries. Our approach effectively bridges this gap by offering clear guidance on object boundaries, thereby elevating the quality of domain adaptation. Our approach undergoes rigorous evaluation on the established unsupervised domain adaptation benchmarks, specifically in adapting SYNTHIA → Cityscapes and SYNTHIA → Mapillary. Experimental results show that the proposed model attains better performance than state-of-the-art methods. The superior performance across different unsupervised domain adaptation scenarios highlights the versatility and robustness of the proposed method.
AB - The generalization capability of unsupervised domain adaptation can mitigate the need for extensive pixel-level annotations to train semantic segmentation networks by training models on synthetic data as a source with computer-generated annotations. Entropy-based adversarial networks are proposed to improve source domain prediction; however, they disregard significant external information, such as edges, which have the potential to identify and distinguish various objects within an image accurately. To address this issue, we introduce a novel approach to domain adaptation, leveraging the synergy of internal and external information within entropy-based adversarial networks. In this approach, we enrich the discriminator network with edge-predicted probability values within this innovative framework to enhance the clarity of class boundaries. Furthermore, we devised a probability-sharing network that integrates diverse information for more effective segmentation. Incorporating object edges addresses a pivotal aspect of unsupervised domain adaptation that has frequently been neglected in the past—the precise delineation of object boundaries. Conventional unsupervised domain adaptation methods usually center around aligning feature distributions and may not explicitly model object boundaries. Our approach effectively bridges this gap by offering clear guidance on object boundaries, thereby elevating the quality of domain adaptation. Our approach undergoes rigorous evaluation on the established unsupervised domain adaptation benchmarks, specifically in adapting SYNTHIA → Cityscapes and SYNTHIA → Mapillary. Experimental results show that the proposed model attains better performance than state-of-the-art methods. The superior performance across different unsupervised domain adaptation scenarios highlights the versatility and robustness of the proposed method.
KW - Depth estimation
KW - Entropy-based adversarial network
KW - Semantic edge detection
KW - Semantic segmentation
KW - Unsupervised domain adaptation
UR - http://www.scopus.com/inward/record.url?scp=85174703196&partnerID=8YFLogxK
U2 - 10.1016/j.eswa.2023.122120
DO - 10.1016/j.eswa.2023.122120
M3 - Article
AN - SCOPUS:85174703196
SN - 0957-4174
VL - 238
JO - Expert Systems with Applications
JF - Expert Systems with Applications
M1 - 122120
ER -