DEEP LEARNING-BASED PERSONALIZED COLOR CORRECTION IN XR

Introduction

Color Vision Deficiency (CVD) affects users’ ability to distinguish certain colors, impacting daily tasks and interactions. Existing solutions, such as color filters or corrective glasses, are often static, non-personalized, and can distort the natural appearance of the environment.

XR enables real-time, adaptive color correction directly within the user’s view. However, designing a system that both improves color distinguishability and preserves natural scene appearance remains challenging. Deep learning offers a promising approach to learn personalized and context-aware color transformations that adapt dynamically to different environments and user needs.

TAsk

• Develop real-time XR techniques for color correction.
• Improve color recognition while maintaining natural visual appearance.

Work

30% Theory, 50% Programming, 20% Writing

Contact

Gwen Qin (gwen.qin@utwente.nl)