91‐4: Temperature Compensation Study of Micro‐LED by Machine Learning

Micro‐LED is a new display technology that offers advantages such as high brightness, high contrast, and low power consumption. However, the performance of Micro‐LED is greatly affected by temperature, leading to color shift easily. To address this issue, this paper proposes a machine learning‐based approach for achieving global compensation, which exhibits advantages including high compensation accuracy, good smoothness, and strong generalization performance. Experimental validation using both measured and simulated data demonstrates that the proposed method achieves a single‐pixel single‐color(R) brightness prediction accuracy of 1% and the brightness prediction of multi‐pixel single‐color(R) also exhibits high accuracy. Consequently, the proposed method enables global brightness compensation and color shift compensation in Micro‐LED.