Facial landmark detection improved by noise adaptive frequency filtering
FAHCD-Net: Frequency-Adaptive Heatmap-Conditional Diffusion Networks for Robust Facial Landmark Detection
Computer Vision and Pattern Recognition
Summary
Facial landmark detection means finding key points on a face, which is important for things like face recognition and animation. The problem is that noise and changes in facial structure can confuse current methods and reduce accuracy. The authors introduce FAHCD-Net, which uses a special step that targets and reduces noisy signals while keeping important face details, producing clearer maps of facial landmarks. They also add a smoothness check to make the results less jumpy, improving performance on tricky images.
What this means in practice
- •For mobile app developers: Develop mobile photo editing apps that accurately detect facial landmarks even in low-quality or noisy images to improve filters and effects.$Commercial implications: Enables more reliable face tracking and effects for consumer apps, enhancing user experience with robust landmark detection.
- •For security system engineers: Implement improved facial landmark detection to strengthen face recognition in security cameras under challenging conditions like poor lighting or distortion.
Authors
Jun Wan, Jiwei Hu, Shengkai Hu, Qilu Zhu
Abstract
Facial Landmark Detection(FLD) is a crucial task in various applications and has achieved significant advancements in recent years. However, current FLD methods still struggle under challenging conditions, where facial structural variations, information loss, and noise interference severely compromise the integrity and accuracy of learned facial features. To address these issues, we propose Frequency-Adaptive Heatmap-Conditional Diffusion Network (FAHCD-Net), which integrates a Frequency-Adaptive Heatmap-Conditional Diffusion (FAHCD) model with a Smoothness Regularization (SR) loss in a cascaded framework. Specifically, the FAHCD model incorporates a Hierarchical Frequency Adaptation (HFA) module designed to suppress redundant high-frequency noise through multi-layer frequency decomposition and adaptive reconstruction, thereby preserving essential facial structures. Additionally, the SR loss is proposed to further mitigate the interference of high-frequency noise and enhance the smoothness of the generated landmark heatmaps. By cascading the FAHCD model with the SR loss, FAHCD-Net effectively leverages both statistical and frequency-based distribution characteristics of the data to progressively generate more accurate landmark heatmaps from noisy inputs. Extensive experiments on popular benchmarks demonstrate the effectiveness and robustness of the proposed method, achieving state-of-the-art performance in FLD tasks under challenging scenarios. The source code is available at https://github.com/HJWKryptonite/FAHCD-Net.