Kichang Yang

Research Interest

I am deeply passionate about developing innovative systems, with a particular interest in Mobile/Edge AI Acceleration and eXtended Reality (XR) Systems. My expertise primarily lies in vision workloads, yet my enthusiasm extends to embracing and mastering new domains. I firmly believe in the power of collaboration to achieve groundbreaking advancements.

Education

Ph.D. in Computer Science & Engineering (2023 ~ Present)
Seoul National University, Seoul, Republic of Korea
Advisor: Prof. Youngki Lee

M.S. in Computer Science & Engineering (2021 ~ 2023)
Seoul National University, Seoul, Republic of Korea
Advisor: Prof. Youngki Lee
Best Thesis Award in Computer Science & Engineering Department

B.S. in Industrial Engineering (2015 ~ 2021)
Seoul National University, Seoul, Republic of Korea
Minor in Computer Science & Engineering
GPA: 4.05/4.30, Summa Cum Laude graduation

Publications

1. NeurIPS

   

   NeurIPS

   VLM in a flash: I/O-Efficient Sparsification of Vision-Language Model via Neuron Chunking

   Kichang Yang^*, Seonjun Kim^*, Minjae Kim, Nairan Zhang, Chi Zhang, Youngki Lee

   In Advances in Neural Information Processing Systems, 2025

   PDF
2. MobiCom

   

   MobiCom

   Logan: Loss-tolerant Video Analytics System

   Kichang Yang, Minkyung Jeong, Juheon Yi, Jingyu Lee, KyungSoo Park, Youngki Lee

   In Proceedings of the 30th Annual International Conference on Mobile Computing and Networking, 2024

   PDF
3. INFOCOM

   

   INFOCOM

   FlexPatch: Fast and Accurate Object Detection for On-device High-Resolution Live Video Analytics

   Kichang Yang, Juheon Yi, Kyungjin Lee, Youngki Lee

   In IEEE INFOCOM 2022 - IEEE Conference on Computer Communications, 2022

   PDF

Experience

Applied Scientist Intern (Jul. 2024 ~ Oct. 2024)
Amazon, California, US
• Efficient On-device Video Understanding with Large Vision Language Models

Data Science Intern (Jun. 2019 ~ Aug. 2019)
SK Hynix, Gyeonggi-do, Republic of Korea
• Deep Learning based DRAM Module Visual Inspection Process (Best Project Award)

Product Manager Intern (Dec. 2018 ~ Feb. 2019)
Educast, Seoul, Republic of Korea
• Linear Algebra for Machine Learning Course Development & Marketing

Republic of Korea Auxiliary Police (Feb. 2017 ~ Nov. 2018)
Korean National Police Agency, Seoul, Republic of Korea

Projects