Learning Deep Representation for Face Alignment with Auxiliary Attributes
Publication in refereed journal


引用次數
替代計量分析
.

其它資訊
摘要In this study, we show that landmark detection or face alignment task is not a single and independent problem. Instead, its robustness can be greatly improved with auxiliary information. Specifically, we jointly optimize landmark detection together with the recognition of heterogeneous but subtly correlated facial attributes, such as gender, expression, and appearance attributes. This is non-trivial since different attribute inference tasks have different learning difficulties and convergence rates. To address this problem, we formulate a novel tasks-constrained deep model, which not only learns the inter-task correlation but also employs dynamic task coefficients to facilitate the optimization convergence when learning multiple complex tasks. Extensive evaluations show that the proposed task-constrained learning (i) outperforms existing face alignment methods, especially in dealing with faces with severe occlusion and pose variation, and (ii) reduces model complexity drastically compared to the state-of-the-art methods based on cascaded deep model.
著者Zhang ZP, Luo P, Loy CC, Tang XO
期刊名稱IEEE Transactions on Pattern Analysis and Machine Intelligence
詳細描述doi>10.1109/TPAMI.2015.2469286.
出版年份2016
月份5
日期1
卷號38
期次5
出版社IEEE COMPUTER SOC
頁次918 - 930
國際標準期刊號0162-8828
電子國際標準期刊號1939-3539
語言英式英語
關鍵詞convolutional network; deep learning; Face Alignment; face landmark detection
Web of Science 學科類別Computer Science; Computer Science, Artificial Intelligence; Engineering; Engineering, Electrical & Electronic

上次更新時間 2020-11-08 於 03:47