Advancing Thin Film Coloration: A Deep Learning Model for Accurate Color Value Predictions

Yun-Jie Jhang^1*, Chia-Hung Chou², Yuan-Fang Lee², Zong-Ying Yang², Hung-Wen Chen^1,2

¹International Intercollegiate Ph.D. Program, National Tsing Hua University, Hsinchu, Taiwan
²Institute of Photonics Technologies, National Tsing Hua University, Hsinchu, Taiwan

* Presenter:Yun-Jie Jhang, email:s111003803@m111.nthu.edu.tw

We propose deep learning models for accurately predicting Lab color values in laser- induced thin film coloration, utilizing both simulated and real-world machining parameters. The models demonstrated impressive performance, achieving ΔE values of 0.77 and 3.55 for simulated and real-world datasets, respectively. The results indicate promising potential for deep learning in enhancing efficiency and precision in predicting laser-induced coloring.

Keywords: Laser Machining Parameters, Deep Learning, Lab Color Space, Laser-induced Coloring, Thin Film Production