Robust Feature Learning Against Noisy Labels

Tsung Ming Tai*, Yun Jie Jhang, Wen Jyi Hwang

*此作品的通信作者

研究成果: 書貢獻/報告類型會議論文篇章

摘要

Supervised learning of deep neural networks heavily relies on large-scale datasets annotated by high-quality labels. In contrast, mislabeled samples can significantly degrade the generalization of models and result in memorizing samples, further learning erroneous associations of data contents to incorrect annotations. To this end, this paper proposes an efficient approach to tackle noisy labels by learning robust feature representation based on unsupervised augmentation restoration and cluster regularization. In addition, progressive self-bootstrapping is introduced to minimize the negative impact of supervision from noisy labels. Our proposed design is generic and flexible in applying to existing classification architectures with minimal overheads. Experimental results show that our proposed method can efficiently and effectively enhance model robustness under severely noisy labels.

原文英語
主出版物標題2023 IEEE International Conference on Image Processing, ICIP 2023 - Proceedings
發行者IEEE Computer Society
頁面2235-2239
頁數5
ISBN(電子)9781728198354
DOIs
出版狀態已發佈 - 2023
事件30th IEEE International Conference on Image Processing, ICIP 2023 - Kuala Lumpur, 马来西亚
持續時間: 2023 10月 82023 10月 11

出版系列

名字Proceedings - International Conference on Image Processing, ICIP
ISSN(列印)1522-4880

會議

會議30th IEEE International Conference on Image Processing, ICIP 2023
國家/地區马来西亚
城市Kuala Lumpur
期間2023/10/082023/10/11

ASJC Scopus subject areas

  • 軟體
  • 電腦視覺和模式識別
  • 訊號處理

指紋

深入研究「Robust Feature Learning Against Noisy Labels」主題。共同形成了獨特的指紋。

引用此