Lab Interview: Vehicle Intelligence Perception Lab.

Our mission is to advance environmental perception, situation understanding, and autonomous control through deep learning and intelligent sensing.

• We develop AI algorithms for autonomous vehicles, focusing on LiDAR–camera fusion, 3D scene understanding, and drivable area detection.

• Our research covers semantic segmentation, object detection & tracking, and 3D map generation for real-time autonomous navigation.

• By integrating deep learning with robotics and sensor engineering, we aim to enhance perception robustness under complex and off-road conditions.

• We work on a broad spectrum of topics — from real-time mapping and localization to collaborative multi-robot perception and SDV-based AI frameworks,
   contributing to safer and smarter autonomous mobility.

VIP-LAB Weekly Seminar(석사)

Title: ACAM-KD: Adaptive and Cooperative Attention Masking for Knowledge Distillation 

Publication / Year: ICCV / 2025

• Contribution

  1. Cross-Attention 기반의 Cooperative Distillation Framework 제안 

  2. 학습 과정에 적응하는 Adaptive Spatial-Channel Masking 

  3. 다양한 Dense Prediction Task에서의 일관된 성능 개선 

VIP-LAB Weekly Seminar(학부생)

Title: MoRe: Class Patch Attention Needs Regularization for Weakly Supervised Semantic Segmentation 

Publication / Year: AAAI / 2025

• Contribution

  1. attention을 graph로 해석 → 필요한 패치만 참조하도록 묵시적 규제 

  2. class-patch attention의 명시적 규제