Kushagra Srivastava1 Damodar Datta Kancharla1 Rizvi Tahereen1 Pradeep Kumar Ramancharla1 Ravi Kiran Sarvadevabhatla1 Harikumar Kandath1
1 Robotics Research Center, IIIT Hyderabad, India
Crack segmentation plays a crucial role in ensuring the struc-tural integrity and seismic safety of civil structures. However, existing crack segmentation algorithms encounter challenges in maintaining ac-curacy with domain shifts across datasets. To address this issue, we pro-pose a novel deep network that employs incremental training with unsu-pervised domain adaptation (UDA) using adversarial learning, without a significant drop in accuracy in the source domain. Our approach leverages an encoder-decoder architecture, consisting of both domain-invariant and domain-specific parameters. The encoder learns shared crack fea-tures across all domains, ensuring robustness to domain variations. Si-multaneously, the decoder’s domain-specific parameters capture domain-specific features unique to each domain. By combining these compo-nents, our model achieves improved crack segmentation performance. Furthermore, we introduce BuildCrack, a new crack dataset comparable to sub-datasets of the well-established CrackSeg9K dataset in terms of image count and crack percentage. We evaluate our proposed approach against state-of-the-art UDA methods using different sub-datasets of CrackSeg9K and our custom dataset. Our experimental results demon-strate a significant improvement in crack segmentation accuracy and generalization across target domains compared to other UDA methods - specifically, an improvement of 0.65 and 2.7 mIoU on source and target domains respectively