@misc{indiciaed21047364347,
  title = {NexViTAD: Few-shot Unsupervised Cross-Domain Defect Detection via Vision Foundation Models and Multi-Task Learning},
  author = {Tianwei Mu and Feiyu Duan and Bo Zhou and Dan Xue and Manhong Huang},
  year = {2025},
  url = {https://arxiv.org/abs/2507.07579},
  note = {Source identifier: 2507.07579}
}