DC 欄位 |
值 |
語言 |
DC.contributor | 資訊工程學系 | zh_TW |
DC.creator | 紀浚鴻 | zh_TW |
DC.creator | Jun-Hong Ji | en_US |
dc.date.accessioned | 2023-8-3T07:39:07Z | |
dc.date.available | 2023-8-3T07:39:07Z | |
dc.date.issued | 2023 | |
dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=110522037 | |
dc.contributor.department | 資訊工程學系 | zh_TW |
DC.description | 國立中央大學 | zh_TW |
DC.description | National Central University | en_US |
dc.description.abstract | 針織布的瑕疵檢測在工業場域中具有重要意義,有助於確保產品質量。
然而,由於不同種類的針織布具有不同的樣態,胚布瑕疵檢測面臨著挑戰。
另外,檢測問題透過深度學習方法通常都取得了不錯的結果,但需要大量標記資料,而在實際情況下這並不容易取得。
此外,現場硬體設備和即時運算的限制要求模型保持輕量化。
因此,我們提出一種適用於少量標記資料的AI方法,以提高訓練效果和實時性能。
同時,我們認識到可解釋性對於工業場域的重要性。
我們的方法不僅關注模型的準確性,還著重於模型可解釋性的部分,增加場域人員對於AI方法的信任度。
總結而言,本研究旨在解決胚布瑕疵檢測中的挑戰,提出一種輕量且可解釋的AI方法,適用於少量標記資料和現場限制。
這將有助於提高瑕疵檢測的效率和準確性,同時增加在工業場域中的可行性和可信度。 | zh_TW |
dc.description.abstract | Defect detection in fabrics is crucial for ensuring product quality in industrial settings.
However, it is challenging due to the diverse patterns of different fabric types.
Deep learning methods have shown promise in defect detection, but they require a large amount of labeled data, which is often difficult to obtain in practical scenarios.
Additionally, the need for lightweight models arises from on-site hardware limitations and real-time processing requirements.
To overcome these challenges, we propose an AI approach that is effective with limited labeled data and offers real-time performance.
We also prioritize interpretability, recognizing its importance in industrial settings.
Our approach not only focuses on model accuracy but also emphasizes interpretability to gain trust from domain experts.
In summary, our research aims to tackle the challenges of defect detection in fabrics by introducing a lightweight and interpretable AI method suitable for limited labeled data and on-site constraints.
This approach will enhance the efficiency and accuracy of defect detection while ensuring feasibility and credibility in industrial environments. | en_US |
DC.subject | 針織布瑕疵檢測 | zh_TW |
DC.subject | 電腦視覺 | zh_TW |
DC.subject | 深度學習 | zh_TW |
DC.subject | 可解釋性AI | zh_TW |
DC.subject | Fabric Defect Detection | en_US |
DC.subject | Computer Vision | en_US |
DC.subject | Deep Learning | en_US |
DC.subject | Explainable AI | en_US |
DC.title | 應用遷移學習於胚布瑕疵檢測 | zh_TW |
dc.language.iso | zh-TW | zh-TW |
DC.title | Applying transfer learning to fabric defect detection | en_US |
DC.type | 博碩士論文 | zh_TW |
DC.type | thesis | en_US |
DC.publisher | National Central University | en_US |