Teaching

My teaching philosophy is centered on building genuine understanding rather than mechanical skill. In AI and computer vision, the gap between copying code and understanding why a model works—or fails—is significant. I try to bridge this gap by grounding every topic in concrete problems and questions before introducing methods.

I believe students learn best when they encounter real obstacles. In workshops, I deliberately include challenging datasets and underperforming baselines so students must diagnose and reason, not just reproduce results. I prioritize teaching research habits—asking the right questions, designing controlled experiments, interpreting results critically—over any specific tool or framework.

Mentoring graduate and undergraduate students in their first research projects is something I find particularly rewarding. I try to help students develop independent scientific judgment, so they can design experiments and evaluate findings without relying on constant guidance.