Multi-task Gaze Estimation Via Unidirectional Convolution

TL;DR

This paper introduces Multitask-Gaze, a lightweight model with novel components like Unidirectional Convolution and attention mechanisms, significantly improving gaze estimation accuracy while reducing computational costs.

Contribution

The paper proposes a new lightweight gaze estimation network, Multitask-Gaze, with innovative modules that enhance performance and efficiency over existing methods.

Findings

Improves gaze estimation accuracy on MPIIFaceGaze and Gaze360 datasets.

Reduces model parameters by 75.5% and FLOPs by 86.88%.

Outperforms state-of-the-art SUGE method in accuracy.

Abstract

Using lightweight models as backbone networks in gaze estimation tasks often results in significant performance degradation. The main reason is that the number of feature channels in lightweight networks is usually small, which makes the model expression ability limited. In order to improve the performance of lightweight models in gaze estimation tasks, a network model named Multitask-Gaze is proposed. The main components of Multitask-Gaze include Unidirectional Convolution (UC), Spatial and Channel Attention (SCA), Global Convolution Module (GCM), and Multi-task Regression Module(MRM). UC not only significantly reduces the number of parameters and FLOPs, but also extends the receptive field and improves the long-distance modeling capability of the model, thereby improving the model performance. SCA highlights gaze-related features and suppresses gaze-irrelevant features. The GCM replaces the pooling layer and avoids the performance degradation due to information loss. MRM improves the accuracy of individual tasks and strengthens the connections between tasks for overall performance improvement. The experimental results show that compared with the State-of-the-art method SUGE, the performance of Multitask-Gaze on MPIIFaceGaze and Gaze360 datasets is improved by 1.71% and 2.75%, respectively, while the number of parameters and FLOPs are significantly reduced by 75.5% and 86.88%.