Abstract

Subject calibration has been demonstrated to improve the accuracy in high-performance eye trackers. However, the true weight of calibration in off-the-shelf eye tracking solutions is still not addressed. In this work, a theoretical framework to measure the effects of calibration in deep learning-based gaze estimation is proposed for low-resolution systems. To this end, features extracted from the synthetic U2Eyes dataset are used in a fully connected network in order to isolate the effect of specific user’s features, such as kappa angles. Then, the impact of system calibration in a real setup employing I2Head dataset images is studied. The obtained results show accuracy improvements over 50%, probing that calibration is a key process also in low-resolution gaze estimation scenarios. Furthermore, we show that after calibration accuracy values close to those obtained by high-resolution systems, in the range of 0.7°, could be theoretically obtained if a careful selection of image features was performed, demonstrating significant room for improvement for off-the-shelf eye tracking systems.

Highlights

  • In recent years, several areas within the world of computer vision have undergone a revolution with the irruption of deep neural networks, systems that have allowed to change the problem-solving approach by shifting the focus towards the quality of the training data used rather than towards the solving method itself

  • This degree of uncertainty is not the same for all types of neural networks: in the case of Convolutional Neural Networks (CNNs) where the input to the network consists of images, the uncertainty stands on the key features of the image that allow the problem to be solved

  • For the analysis of the results of the experiments regarding the theoretical framework, we present in Figure 7 the boxplot distributions of the angular error for the different users and experiments configuration

Read more

Summary

Introduction

Several areas within the world of computer vision have undergone a revolution with the irruption of deep neural networks, systems that have allowed to change the problem-solving approach by shifting the focus towards the quality of the training data used rather than towards the solving method itself. This, from a research point of view, can be negative as it hinders the establishment of a theoretical framework, which defines what elements should appear as input data to make the problem solvable. This degree of uncertainty is not the same for all types of neural networks: in the case of Convolutional Neural Networks (CNNs) where the input to the network consists of images, the uncertainty stands on the key features of the image that allow the problem to be solved.

Objectives
Methods
Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call