Joonhyun Jeong

AI Research Engineer

About Me

Best Developer Enjoying Life with AI.
My Research Interests are

  • Computer Vision
  • Multi-Modal Vision-Language Models
  • Object Detection
  • Model Compression
  • Neural Architecture Search
  • Data Augmentation

Experience

Codigm (Goorm)

Mar 2018 - Jun 2018

https://goorm.co/

Web Front & Backend Engineer / Intern

Sep 2019 - Feb 2020

https://clova.ai/ko

AI Model Research & Develop for Face Detection task / Intern (Face team)

May 2020 - Current

https://clova.ai/ko

AI Research Engineer / Full-Timer in Image Vision team under the supervision of Youngjoon Yoo.

Education

Kyung-Hee University

Bachelor Degree, Department of Computer Science and Engineering

Mar 2015 - Feb 2019

Kyung-Hee University

Department of Computer Science and Engineering

Mar 2019 - Aug 2019

Machine Learning and Visual Computing Laboratory (MLVC) Lab under the supervision of Prof. Sung-Ho Bae.

Korea Advanced Institute of Science and Technology (KAIST)

Master Degree, Kim Jaechul Graduate School of AI (GSAI)

Sep 2022 - Current

Machine Learning and Intelligence LAB (MLI) Lab under the supervision of Prof. Eunho Yang.

Publications

Hijacking Context in Large Multi-modal Models

ICLRW 2024 (R2-FM Workshop)

Joonhyun Jeong

[Paper]

ProxyDet: Synthesizing Proxy Novel Classes via Classwise Mixup for Open Vocabulary Object Detection

AAAI 2024

Joonhyun Jeong, Geondo Park, Jayeon Yoo, Hyungsik Jung, Heesu Kim

[Paper]

EResFD: Rediscovery of the Effectiveness of Standard Convolution for Lightweight Face Detection

WACV 2024

Joonhyun Jeong, Beomyoung Kim, Joonsang Yu, Youngjoon Yoo

[Paper]

GeNAS: Neural Architecture Search with Better Generalization

IJCAI 2023

Joonhyun Jeong, Joonsang Yu, Geondo Park, Dongyoon Han, Youngjoon Yoo

[Paper]

The Devil is in the Points: Weakly Semi-Supervised Instance Segmentation via Point-Guided Mask Representation

CVPR 2023

Beomyoung Kim, Joonhyun Jeong, Dongyoon Han, Sung Ju Hwang

[Paper]

Neural Architecture Search with Loss Flatness-aware Measure

ICMLW 2022 (DyNN Workshop)

Joonhyun Jeong, Joonsang Yu, Dongyoon Han, YoungJoon Yoo

[Paper]

Dataset Condensation via Efficient Synthetic-Data Parameterization

ICML 2022

Jang-Hyun Kim, Jinuk Kim, Seong Joon Oh, Sangdoo Yun, Hwanjun Song, Joonhyun Jeong, Jung-Woo Ha, Hyun Oh Song

[Paper]

Observations on K-image Expansion of Image-Mixing Augmentation for Classification

IEEE ACCESS 2023

Joonhyun Jeong, Sungmin Cha, Youngjoon Yoo, Sangdoo Yun, Taesup Moon, Jongwon Choi

[Paper]

Bayesian Perspective on Visual Data Augmentation for Efficient Utilization of Subsampled Data

ICLRW 2021 (Synthetic Data Generation Workshop)

Joonhyun Jeong, Sungmin Cha, Youngjoon Yoo, Sangdoo Yun, Jongwon Choi

[Paper]

An Inter-Layer Weight Prediction and Quantization for Deep Neural Networks based on a Smoothly Varying Weight Hypothesis

Arxiv 2019

Kang-Ho Lee, JoonHyun Jeong, Sung-Ho Bae

[Paper]

A New Pointwise Convolution in Deep Neural Networks Through Extremely Fast and Non Parametric Transforms

IEEE ACCESS 2022

Joonhyun Jeong, Incheon Cho, Eunseop Shin, Sung-Ho Bae

[Paper]

고속 이산 코사인 변환을 이용한 새로운 경량화 콘볼루션 신경망

2019 한국컴퓨터종합학술대회 (KCC2019), 우수논문상

Joonhyun Jeong, Sung-Ho Bae

[Paper]

Projects

Automatic Authentication System Based on Face Recognition.

I took a role of modeling part for Real-Time Face Detection & Facial Landmark Detection, aimed for extremely fast inference without accuracy performance degradation.

Talks

Invited Talk from Naver Connect Foundation

[Link]

Invited talk from Naver Connect Foundation, as a mentor for undergraduate students.

Invited Lecture in Kyung-Hee University

[Video]

Invited lecture for career search seminar in Kyung-Hee University.

KCC 2019 Oral Talk

[Youtube]

Invted talk for introducing the paper “A New Pointwise Convolution in Deep Neural Networks Through Extremely Fast and Non Parametric Transforms”