Lite-HRNet Plus: Fast and Accurate Facial Landmark Detection

TL;DR

Lite-HRNet Plus introduces a novel architecture for facial landmark detection that enhances accuracy and reduces computational cost, making it suitable for real-time applications.

Contribution

It proposes a new fusion block with channel attention and a lightweight output module, improving upon Lite-HRNet for faster and more accurate facial landmark detection.

Findings

Achieved state-of-the-art accuracy on facial landmark datasets.

Reduced computational complexity to around 10M FLOPs.

Improved fusion efficiency with a novel channel attention-based block.

Abstract

Facial landmark detection is an essential technology for driver status tracking and has been in demand for real-time estimations. As a landmark coordinate prediction, heatmap-based methods are known to achieve a high accuracy, and Lite-HRNet can achieve a fast estimation. However, with Lite-HRNet, the problem of a heavy computational cost of the fusion block, which connects feature maps with different resolutions, has yet to be solved. In addition, the strong output module used in HRNetV2 is not applied to Lite-HRNet. Given these problems, we propose a novel architecture called Lite-HRNet Plus. Lite-HRNet Plus achieves two improvements: a novel fusion block based on a channel attention and a novel output module with less computational intensity using multi-resolution feature maps. Through experiments conducted on two facial landmark datasets, we confirmed that Lite-HRNet Plus further improved the accuracy in comparison with conventional methods, and achieved a state-of-the-art accuracy with a computational complexity with the range of 10M FLOPs.