Heatmap Regression without Soft-Argmax for Facial Landmark Detection
This work addresses a bottleneck in computer vision for applications such as head pose estimation and face swapping, offering an incremental improvement over existing heatmap regression methods.
The paper tackles the problem of facial landmark detection by proposing an alternative to the widely used Soft-argmax method, achieving state-of-the-art performance on benchmarks like WFLW, COFW, and 300W with 2.2x faster training convergence.
Facial landmark detection is an important task in computer vision with numerous applications, such as head pose estimation, expression analysis, face swapping, etc. Heatmap regression-based methods have been widely used to achieve state-of-the-art results in this task. These methods involve computing the argmax over the heatmaps to predict a landmark. Since argmax is not differentiable, these methods use a differentiable approximation, Soft-argmax, to enable end-to-end training on deep-nets. In this work, we revisit this long-standing choice of using Soft-argmax and demonstrate that it is not the only way to achieve strong performance. Instead, we propose an alternative training objective based on the classic structured prediction framework. Empirically, our method achieves state-of-the-art performance on three facial landmark benchmarks (WFLW, COFW, and 300W), converging 2.2x faster during training while maintaining better/competitive accuracy. Our code is available here: https://github.com/ca-joe-yang/regression-without-softarg.