GazeFollowTR: A Method of Gaze Following with Reborn Mechanism

摘要

Gaze following is the task of estimating where an observer islooking inside a scene. Both the observer and scene information must belearned to determine the gaze directions and gaze points. Recently, manyexisting works have only focused on scenes or observers. In contrast,revealed frameworks for gaze following are limited. In this paper, a gazefollowing method using a hybrid transformer is proposed. Based on theconventional method (GazeFollow), we conduct three developments. First,a hybrid transformer is applied for learning head images and gaze positions.Second, the pinball loss function is utilized to control the gaze point error.Finally, a novel ReLU layer with the reborn mechanism (reborn ReLU) isconducted to replace traditional ReLU layers in different network stages.To test the performance of our developments, we train our developedframework with the DL Gaze dataset and evaluate the model on ourcollected set. Through our experimental results, it can be proven that ourframework can achieve outperformance over our referred methods.