The Perception and Intelligence Laboratory (KNU-PIL) studies how artificial intelligence systems perceive the real world and transform sensory information into reliable and generalizable intelligence.

Our research broadly covers

• Perception and representation learning in real-world environments

• Learning from multiple-modalities (e.g. image, text), heterogeneous, or imperfect inputs

• Robust and reliable decision-making under practical constraints

• Data-efficient machine learning methods

Our goal

• To understand why AI systems succeed or fail in real-world scenarios

• To design learning methods that remain reliable under realistic conditions

• To bridge perception and intelligence for practical AI applications