Aigerim Saken

Here’s a bonus photo of beautiful Prague, where the conference is taking place

Instead of classifying each pixel, our approach teaches an agent to identify the center of each grain based on structural patterns, even when images are noisy or ambiguous.

Had a great time presenting my research at the 19th Learning and Intelligent Optimization (LION19) conference today!

I showcased our work on using reinforcement learning to improve grain image segmentation—a key step in materials analysis. 1/n