本論文針對TFT面板瑕疵分類問題，探討以半監督式域自適應學習的方式，以達到縮短TFT Defect 分類網路在面對相同類別但新產品樣本資料的標記成本與訓練時間。論文的主要研究目標有二：第一，利用非監督式的網路架構，設計一個可拉開不同類別樣本間特徵距離的作法，以提高面板瑕疵分類的效果；第二，針對網路損失函數的設計進行分析與改善，以在維持原有網路架構的分類成效下，更適用於TFT Defect 分類問題。在實驗分析上，我們實際測試於包含5個產品的TFT Defect Dataset，並探討在不同比例的標註資料、不同參數設計、不同網路損失函數等對於分類成效的影響。實驗結果顯示，在只有少量Target Label的情況下，域自適應的深度學習網路架構的確可以成功應用在 TFT Defect 分類問題上，取得接近甚至比重新訓練更好的分類結果，有效地減少整體的模型訓練時間與資源成本。

This thesis focuses on thin-film-transistor (TFT) defect classification via semi-supervised domain adaptation. With the rapidly evolving TFT manufacturing process, automatic defect classification has become indispensable for TFT-LCD industry. Although deep-learning based methods have achieved remarkable performance for many classification tasks, the success of model training heavily relies on large-scaled and accurately-labeled datasets. In this thesis, we study and experiment using semi-supervised domain adaptation to facilitate the model trained on fully-labeled source domain into the target domain with few labeled data. Our goal is two-fold: to enlarge the inter-class feature discriminability alone with the model adaptation, and to study and compare different hyperparameter settings and loss terms for the TFT defect classification. Experimental results on the TFT Defect Dataset verify the effectiveness of our study and demonstrate its applicability to real-world problems.

Abstract .......................................... III
Contents .......................................... IV
List of Figures ................................... VI
List of Tables .................................... IX
1 Introduction ..................................... 1
1.1 Motivation ..................................... 1
1.2 Outline ........................................ 2
2 Related Work ..................................... 4
2.1 TFTDefect Detection ............................ 4
2.2 Domain Adaptation .............................. 8
2.3 Models of Unsupervised Domain Adaptation ...... 14
3 Methodology ..................................... 17
3.1 ProblemStatement .............................. 17
3.2 Motivation .................................... 19
3.3 Divergence-based Domain Adaptation Model....... 19
3.4 ContrastiveAdaptationNetworkBaseline........... 21
4 Experiments ..................................... 23
4.1 ExperimentalSetting ........................... 23
4.2 AblationStudy ................................. 27
4.3 Experimental Results and Comparison ........... 29
5 Conclusion ...................................... 32
Bibliography ...................................... 33

1. Mingsheng Long, Yue Cao, Jianmin Wang and Michael Jordan. Learning trans- ferable features with deep adaptation networks.In Proceedings of the 32nd In- ternational Conference on Machine Learning (2015).

2. Mingsheng Long, Han Zhu, Jianmin Wang and Michael Jordan. Deep transfer learning with joint adaptation networks. In Proceedings of the 34th Interna- tional Conference on Machine Learning (2017).

3. Eric Tzeng, Judy Ho↵man, Ning Zhang, Kate Saenko and Trevor Darrell. Deep domain confusion: Maximizing for domain invariance. arXiv:1412.3474 [cs.CV] (2014).

4. Guoliang Kang, Lu Jiang, Yi Yang and Alexander G. Hauptmann. Contrastive Adaptation Network for Unsupervised Domain Adaptation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(2019), pp. 4893-4902.

5. Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pascal Germain, Hugo Larochelle, Fran ̧cois Laviolette, Mario Marchand and Victor Lempitsky. Domain-adversarial training of neural networks. Journal of Machine Learning Research 17, (2016), Vol. 17, Issue 1, pp. 2096-2030.

6. Swami Sankaranarayanan, Yogesh Balaji, Carlos D. Castillo and Rama Chel- lappa. Generate to adapt: Aligning domains using generative adversarial net-
33works. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)(2018), pp. 8503-8512.

7. Eric Tzeng, Judy Ho↵man, Kate Saenko and Trevor Darrell. Adversarial dis- criminative domain adaptation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, (CVPR)(2017), pp. 7167-7176.

8. Konstantinos Bousmalis, George Trigeorgis, Nathan Silberman, Dilip Krishnan and Dumitru Erhan. Domain separation networks, arXiv:1608.06019, (2016).

9. Muhammad Ghifary, W. Bastiaan Kleijn, Mengjie Zhang, David Balduzzi and Wen Li. Deep reconstruction-classification networks for unsupervised domain adaptation. In Proceedings of the European Conference on Computer Vision (ECCV)(2016).

10. Garrett Wilson and Diane J. Cook. A Survey of Unsupervised Deep Do- main Adaptation. In ACM Transactions on Intelligent Systems and Tech- nology Volume 11, Issue 5, (September 2020), Article No.: 51, pp. 1–46, https://doi.org/10.1145/3400066

11. Christian Szegedy, Sergey Io↵e, Vincent Vanhoucke and Alexander A Alemi. Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning. In Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence (AAAI)(2017).

12. Ruifang Ye1, Chia-Sheng Pan, Ming Chang and Qing Yu. Intelligent defect classification system based on deep learning. In Advances in Mechanical Engi- neering, (2018), Vol. 10(3) 1–7 https://doi.org/
10.1177/1687814018766682

13. Sinno Jialin Pan and Qiang Yang. A Survey on Transfer Learning. In IEEE Transactions on Knowledge and Data Engineering 22, (Oct 2010), No.: 10

14. Dino Sejdinovic, Bharath Sriperumbudur, Arthur Gretton and Kenji Fuku- mizu. Equivalence of distance-based and rkhs-based statis- tics in hypothe- sis testing. In The Annals of Statistics, (October 2013), Vol. 41, No.: 5, pp. 2263–2291.

15. Eric Tzeng, Judy Ho↵man, Trevor Darrell and Kate Saenko. Simultaneous Deep Transfer Across Domains and Tasks. In Proceedings of the IEEE Inter- national Conference on Computer Vision, (ICCV)(2015), pp. 4068-4076.

16. Yaroslav Ganin and Victor Lempitsky. Unsupervised Domain Adaptation by Backpropagation. In Proceedings of the 32nd International Conference on Ma- chine Learning, (2015), PMLR 37:1180-1189.

17. Kuniaki Saito, Yoshitaka Ushiku and Tatsuya Harada. Asymmetric Tri- training for Unsupervised Domain Adaptation. In Proceedings of the 34th In- ternational Conference on Machine Learning, (2017), PMLR 70:2988-2997.

18. Yu Zhang and Qiang Yang. A Survey on Multi-Task Learning. In IEEE Trans- actions on Knowledge and Data Engineering, (2017).

19. Rajat Raina, AlexisBattle, HonglakLee, BenjaminPacker and Andrew Y. Ng. Self-taught learning: transfer learning from unlabeled data. In Proceedings of the 24th international conference on Machine learning, (June 2007), pp 759–766, https://doi.org/10.1145/1273496.1273592

20. Andrew Arnold, Ramesh Nallapati and William W. Cohen. A Comparative Study of Methods for Transductive Transfer Learning. In Seventh IEEE Inter- national Conference on Data Mining Workshops, (ICDMW 2007).

21. Wenyuan Dai, Qiang Yang, Gui-Rong Xue and Yong Yu. Self-Taught Cluster- ing. In Proceedings of the 25th international conference on Machine learning, (July 2008), pp. 200–207, https://doi.org/10.1145/1390156.1390182

22. Zheng Wang, Yangqiu Song, and Changshui Zhang. Transferred Dimensional- ity Reduction. In European Conf. Machine Learning and Knowledge Discovery in Databases, (ECML/PKDD)(2008), pp. 550-565.

23. Mingsheng Long, Yue Cao, Jianmin Wang and Michael Jordan. Learning transferable features with deep adaptation networks. In Proceedings of the 32nd International Conference on Machine Learning, (2015), PMLR 37:97-105.