Github| LinkedIn | CV

E-mail: higok18@pusan.ac.kr

Changwoo Baek

I'm a master's student working on hardware-efficient Multimodal LLMs in the Department of Electronics Engineering at Pusan National University.

My goal is to make these models lightweight, fast, and reliable enough to run on edge devices (e.g. mobile, robotics). I’m particularly interested in uncovering how MLLMs work internally, and in using that understanding to derive general principles rather than task-specific heuristics.

Interests:

Efficient (M)LLMs
Visual Token Pruning
KV Cache Compression
Robot Vision
Knowledge Distillation

Efficiency Research

Visual Token Pruning [MLLM]

Knowledge Distillation [VLM]

KV Cache Compression [LLM]

Efficient Architecture [Robot Vision]

News

[2026/01] Our paper AgilePruner is accepted at ICLR 2026

[2025/11] Our paper is accepted at NeurIPS 2025 ER Workshop

[2025/06] Our paper is accepted at IEEE Signal Processing Letters

Educations

Pusan National University, Busan, Republic of Korea (Mar. 2025 - )
M.S. Department of Electronics Engineering
Advisor: Prof. Kyeongbo Kong

Pusan National University, Busan, Republic of Korea (Mar. 2019 - Feb. 2025)
B.S. Department of Electronics Engineering

Publications

(* equal contribution, † corresponding author)

Under Review

Anonymized Title (Spatial-Aware 3D Token Pruning [3D VLM])

C. Baek, K. Kong†

Anonymized Title (Query-agnostic KV Cahce Compression [LLM])

C. Baek, K. Kong†

Anonymized Title (Robot Vision)

C. Baek*, T. Kim*, W. Lee*, J. Son, S. Lee, J. Kim, K. Kong†

International Journal

[J1] Unsupervised Knowledge Distillation via Local Representations for Vision-Language Models

C. Baek, J. Song, K. Kong†

IEEE Signal Processing Letters (SPL), Aug. 2025

International conference

[C3] AgilePruner: An Empirical Study of Attention and Diversity for Adaptive Visual Token Pruning in Large Vision-Language Models

C. Baek*, J. Song*, S. Kim*, K. Kong†

ICLR 2026, Apr. 2026 [Github]

*Initial version was presented at NeruIPS Workshop on Efficient Reasoning, Dec. 2025

[C2] Efficient Token Pruning for Vision-Language Models in Multi-turn Scenarios

C. Baek*, S. Kim*, J. Song*, K. Kong†

IEEE Asia Pacific Conference on Circuits and Systems (APCCAS), Oct. 2025

[C1] Low-Light Face Recognition for Mobile Robots

C. Baek, J. Song, K. Kong†

ITC-CSCC, Jul. 2025 (Special session)

Domestic Journal

[J2] Visual Token Pruning Based on Entropy Dynamics in Vision Encoders of Multimodal LLMs

C. Baek*, S. Kim*, J. Song*, K. Baek, K. Kong†

The Journal of the Korean Institute of Broadcast and Media Engineers (KIBME) Letters, 2026

[J1] Deep Learning-Based Face Recognition through Low-Light Enhancement

C. Baek, K. Kong†

IEMEK Journal of Embedded Systems and Applications, Oct. 2024

Projects

Developed vision-language model-based technologies for visual perception and situation awareness in home robot systems

Collaborating with LG Electronics, Seoul, Korea (June. 2025 - )

Development of Anomaly and Environment Detection Technology for Home Robot Applications

Collaborating with LG Electronics, Seoul, Korea (June. 2025 - Nov. 2025)

Development of Nighttime Human Detection and Daytime Face Recognition Enhancement for Home Robot Environments

Collaborated with LG Electronics, Seoul, Korea (Dec. 2023 - Dec. 2024)

Awards

Outstanding Award, Industrial AI Convergence Club

Industrial AI Specialized Talent Development Program, Pusan National University, 2024

Extra

Teaching Experiences

[EE1600575] C Programming, Teaching Assistant

Dept. of EE, Pusan National University, Spring 2025

[EE3100695] Digital Circuits Laboratory, Teaching Assistant

Dept. of EE, Pusan National University, Fall 2025

Scholarship

Graduate Research Encouragement Scholarship

National Research Foundation of Korea (NRF), 2025

Skills

Python, Pytorch, C, C++, ROS2

Page updated

Google Sites

Report abuse