Have a personal or library account? Click to login
Three-Dimensional Line-of-Sight Estimation Based on RTACM-Net and Vision Transformer Cover

Three-Dimensional Line-of-Sight Estimation Based on RTACM-Net and Vision Transformer

By: Tingjuan Sang and  Wuqi Gao  
Open Access
|Dec 2025

Figures & Tables

Figure 1.

Model Framework
Model Framework

Figure 2.

The structures of TACmix-BTNK1 and TACmix-BTNK2
The structures of TACmix-BTNK1 and TACmix-BTNK2

Figure 3.

ACmix module structure
ACmix module structure

Figure 4.

The TeLU activation function and its derivatives
The TeLU activation function and its derivatives

COMPARISON OF EXPERIMENTAL RESULTS ON MPIIFACEGAZE (°)

modelAverage angular error
L2CS-Net[7]3.92
CA-Net[4]4.10
GazeTR[6]4.00
MPIIGaze[2]5.40
AGE-Net[5]4.09
Dilated-Net[3]4.10
Res-Swin-Ge[8]3.75
RT-Gene[9]4.36
Ours model3.72

COMPARISON OF EXPERIMENTAL RESULTS ON GAZE360 (°)

modelAverage angular error
Full-Face[10]14.99
CA-Net[4]11.20
GazeTR[6]10.67
Gaze360[13]11.04
Dilated-Net[3]13.73
Bot2L-Net[12]11.53
RT-Gene[9]12.26
Ours model10.46
Language: English
Page range: 12 - 21
Published on: Dec 31, 2025
In partnership with: Paradigm Publishing Services
Publication frequency: 4 issues per year

© 2025 Tingjuan Sang, Wuqi Gao, published by Xi’an Technological University
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.