Joo Chan Lee

I am a Ph.D. student in the department of Artificial Intelligence at Sungkyunkwan University, advised by Jong Hwan Ko and Eunbyung Park. My research interest lies in the areas of computer vision, graphics, and machine learning. Currently, I am interested in designing efficient neural fields architecture.

Email  /  Scholar  /  LinkedIn  /  Github

profile photo


Representative papers are highlighted.

Continuous Memory Representation for Anomaly Detection
Joo Chan Lee*, Taejune Kim*, Eunbyung Park, Simon S. Woo, Jong Hwan Ko
arXiv:2402.18293, 2024

A novel approach to learning normal representation in continuous feature space for anomaly detection.

Compact 3D Gaussian Representation for Radiance Field
Joo Chan Lee, Daniel Rho, Xiangyu Sun, Jong Hwan Ko, Eunbyung Park
CVPR, 2024 (Highlight)

A comprehensive framework for 3D scene representation, achieving high performance, fast training, compactness, and real-time rendering.

Coordinate-Aware Modulation for Neural Fields
Joo Chan Lee, Daniel Rho, Seungtae Nam, Jong Hwan Ko, Eunbyung Park
ICLR, 2024 (Spotlight)

Injecting spectral bias-free grid representations into the intermediate features of the MLP achieves high performance with compactness.

FFNeRV: Flow-Guided Frame-Wise Neural Representations for Videos
Joo Chan Lee, Daniel Rho, Jong Hwan Ko, Eunbyung Park
ACM MM, 2023

Incorporating flow information into frame-wise representations to exploit the temporal redundancy across the frames in videos.

KERNTROL: Kernel Shape Control Toward Ultimate Memory Utilization for In-Memory Convolutional Weight Mapping
IEEE TCAS-I, 2024 [Paper]
Kernel Shape Control for Row-Efficient Convolution on Processing-In-Memory Arrays ICCAD, 2023 [Paper] [Code]
Johnny Rhe, Kang Eun Jeon, Joo Chan Lee, Seongmoon Jeong, Jong Hwan Ko

A novel pruning method for a Processing-In-Memory hardware.

Masked Wavelet Representation for Compact Neural Radiance Fields
Daniel Rho*, Byeonghyeon Lee*, Seungtae Nam, Joo Chan Lee, Jong Hwan Ko, Eunbyung Park
CVPR, 2023

Using the wavelet transform with learnable masking for compact grid-based neural radiance fields.

A Reconfigurable Neural Architecture for Edge–Cloud Collaborative Real-Time Object Detection
Joo Chan Lee, Yongwoo Kim, SungTae Moon, Jong Hwan Ko
IEEE Internet of Things Journal, 2022

A single-weight reconfigurable object detector for collaborative intelligence.

Scalable Color Quantization for Task-Centric Image Compression
Jae Hyun Park, Sang Hoon Kim, Joo Chan Lee, Jong Hwan Ko
ACM TOMM, 2022

Images with variable color space sizes can be extracted from a master image generated by a single DNN model.

A Splittable DNN-Based Object Detector for Edge-Cloud Collaborative Real-Time Video Inference
Joo Chan Lee, Yongwoo Kim, SungTae Moon, Jong Hwan Ko
AVSS, 2021

A splittable object detector for real-time collaborative inference.

Robust detection of small and dense objects in images from autonomous aerial vehicles
Joo Chan Lee, JeongYeop Yoo, Yongwoo Kim, SungTae Moon, Jong Hwan Ko
EL, 2021

Technical report for high performance small object detection.

VisDrone-DET2020: The Vision Meets Drone Object Detection in Image Challenge Results
Challenge Participants
ECCV Workshops, 2020

VisDrone Challenge 1st Place Winner.

Design and source code from Jon Barron's website.