Adaptive Wing Loss for Robust Face Alignment via Heatmap Regression

Abstract

Heatmap regression with a deep network has become one of the mainstream approaches to localize facial landmarks. However, the loss function for heatmap regression is rarely studied. In this paper, we analyze the ideal loss function properties for heatmap regression in face alignment problems. Then we propose a novel loss function, named Adaptive Wing loss, that is able to adapt its shape to different types of ground truth heatmap pixels. This adaptability penalizes loss more on foreground pixels while less on background pixels. To address the imbalance between foreground and background pixels, we also propose Weighted Loss Map, which assigns high weights on foreground and difficult background pixels to help training process focus more on pixels that are crucial to landmark localization. To further improve face alignment accuracy, we introduce boundary prediction and CoordConv with boundary coordinates. Extensive experiments on different benchmarks, including COFW, 300W and WFLW, show our approach outperforms the state-of-the-art by a significant margin on various evaluation metrics. Besides, the Adaptive Wing loss also helps other heatmap regression tasks. Code will be made publicly available at https://github.com/protossw512/AdaptiveWingLoss.

arXiv comment[v2] Camera-ready version for ICCV 2019. [v3] Corrected AUC(fr10%) on table 2

References in corpus (5)

• Learning Deep Representation for Face Alignment with Auxiliary Attributes
• Jointly Optimize Data Augmentation and Network Training: Adversarial Data Augmentation in Human Pose Estimation
• Joint Multi-view Face Alignment in the Wild
• Supervision-by-Registration: An Unsupervised Approach to Improve the Precision of Facial Landmark Detectors
• GoDP: Globally optimized dual pathway system for facial landmark localization in-the-wild

Cited by in corpus (6)

• PropagationNet: Propagate Points to Curve to Learn Structure Information
• Spine intervertebral disc labeling using a fully convolutional redundant counting model
• ADNet: Leveraging Error-Bias Towards Normal Direction in Face Alignment
• Think about boundary: Fusing multi-level boundary information for landmark heatmap regression
• Robust Facial Landmark Detection by Cross-order Cross-semantic Deep Network
• Delving into Rectifiers in Style-Based Image Translation