Yooseung Wang

Yooseung Wang (왕유승)

I am a researcher at Mobility Platform Research Center, Korea Electronics Techonlogy Institute (KETI).

Previously, I worked as a research officer (8 EE students selected nationwide) at Defense Artifical Intelligence Center, Agency for Defense Development (ADD). I received my M.S. in Electrical Engineering from Korea Advanced Institute of Science and Technology (KAIST), where I was advised by Prof. Changick Kim, and my B. S. in Electrical and Computer Engineering from Ulsan National Institute of Science and Technology (UNIST).

I am a recipient of 2024 1st Presidential Science Scholarship for Graduate School (9 EE students selected nationwide), 2023 Kwanjeong Scholarship (the sole recipient from KAIST EE/CS), Kwon Yeong-Se Scholarship (Top Academic Entrance Scholarship in KAIST EE (1st/2nd Place)), and graduated as the Valedictorian of the UNIST ECE, Class of 2019.

My research interests include vision-based on-device AI, autonomous driving, and multimodal learning.

Email: yswang@keti.re.kr; yswang@kaist.ac.kr

[CV] [Google scholar]

Education

2022.08 - 2024.08 M.S., School of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), South Korea

- Computation Intelligence Lab, Advisor : Changick Kim, GPA: 4.25/4.3

- Thesis : Subdivided Mask Dispersion Network in Semi-Supervised Semantic Segmentation

2015.03 - 2019.02 B.S. Dep. of Electrical and Computer Engineering, Ulsan National Institute of Science and Technology (UNIST), South Korea
- Valedictorian in ECE (Ranked 1/131), Summa Cum Laude

-2018-2019, Undergraduate Research Intern, CSAIL, Advisor: Jaesik Choi (Was at UNIST, moved to KAIST AI Graduate School)

-2018, Uzbekistan, TUIT(타슈켄트정보통신대학교), ALgorithm Lecturer (English), fully funded by Worlds Friends IT

-2017, Hong Kong, HKUST(홍콩과학기술대학교), Exchange Student, fully funded by government scholarship

-2016, Singapore, NTU(난양공과대학교), Singapore, Participated Global Startup Acceleration Program, fully funded by NIA (winning the Startup Competition)

Research Experience

(2024.09.01~) Korea Electronics Techology Institute (KETI), Mobility Platform Research Center, Researcher

(2019.06-2022.05) Agency for Defense Development (ADD), Defence AI Center (DAIC), Research Officer (Mandatory military Service)

Projects: Autonomous driving car perception [J1, J2, J4, D1, D2, PT1, PT2] (News) / occluded object detection / face recognition

(2018.06-2018.08) Electronics and Telecommunications Research Institute (ETRI),Visual Intelligence Lab, Undergraduate Researcher

Action recognition, object tracking, semantic segmentation

(2018.12 - 2019.02, 2022.06 - 2022.08) Satrec Initiative Analytics (SIA), Image Processing Team, Research Intern

3rd Place, 2022 CVPRW SpaceNet 8 Challenge (link, See sianalytics)
Beat the winner @ 2018 CVPRW DeepGlobe Challenge [GRSL'21]

International Publications

(*bold: first-authored papers)

Machine Perception : Segmentation [ICPR'20, SPL'21a, PRL'22b, PRL'23, PR'24, ACCESS'24], Detection [ICCAS'20, PRL'22b], EO/Lidar Sensor Fusion [TITS'22]
Human Recognition : Human Parsing [PR'24], Face Recognition [ACCESS'22], Group Action Recognition [ECCV'24a, ECCV'24b]
Data Minimization : Few-Shot Learning [SPL'21a], Zero-Shot Learning [PRL'22b], Semi-Supervised Learning [PRL'24, ACCESS'24]
Network Minimization : Network Pruning [SPL'21b], Light-Weighted Network [ICPR'20, GRSL'22]
Multimodal Learning : Prompt-based Learning [VCIP'23, ICASSP'24]

2022~2024 (KAIST)

[C7] Spatio-Temporal Proximity-Aware Dual-Path Model for Panoramic Activity Recognition

S. Lee, Yooseung Wang, S. Woo, Changick Kim

ECCV 2024

[C5] Flow-Assisted Motion Learning Network for Weakly-Supervised Group Activity Recognition

M. A. Nugroho, S. Woo, S. Lee, J. Park, Yooseung Wang, D. Kim, Changick Kim

ECCV 2024

[J10] FCGNet: Foreground and Class Guidd Network for Human Parsing

J. Jang, Yooseung Wang, Changick Kim

Pattern Recognition (SCIE, IF: 8.5), 2023 [Paper]

[J9] CAFS: Class Adative Framework for Semi-Supervised Semantic Segmentation
Jingi Ju*, H. Noh*, Yooseung Wang*, M. Seo, D. Choi

IEEE ACCESS (SCIE, IF: 3.9), 2024 [Paper]

[J8] Subdivided Mask Dispersion Framework for Semi-Supervised Semantic Segmentation

Yooseung Wang, J. Jang, Changick Kim

Pattern Recognition Letters (SCIE, IF: 5.1), 2024 [Preprint, Paper]

[C4] Toward Robust Multimodal Prompting with Missing Modalities

J. Jang, Yooseung Wang, Changick Kim

ICASSP 2024 [Paper]

[C3] Noise-Augmented Missing Modality Aware Prompt based Learning for Robust Visual Recognition

Yooseung Wang, J. Jang, Changick Kim

IEEE VCIP 2023 [Paper, Poster, Conference]

2020-2022 (Defense AI Center, ADD)

[J7] Occlusion-Aware Spatial Attention Transformer for Partially Occluded Object Recognition

J. Heo, Yooseung Wang, J. Park

Pattern Recognition Letters (SCIE, IF: 5.1), 2022 [Paper]

[J6] Angular Margin Mining Softmax for Face Recognition

J. Lee, Yooseung Wang, S. Cho

IEEE ACCESS (SCIE, IF:3.5), 2022 [Paper]

[J5] Zero-shot Semantic Segmentation via Spatial and Multi-Scale Aware Visual Class Embedding

S. Cha*, Yooseung Wang*

Pattern Recognition Letters (SCIE, IF:5.1), 2022 [arXiv, Paper, News]

[J4] Drivable Dirt Road Region Identification Using Image and Point Cloud Semantic Segmentation Fusion

J. Park, H. Yoo, Yooseung Wang

IEEE Transactions on Intelligent Transportation System (SCIE, IF:9.6), 2022 [Paper] [News]

[J3] NL-LinkNet : Toward Lighter but More Accurate Road Extraction with Non-Local Operations

Yooseung Wang, J. Seo, T. Jeon

IEEE Geoscience and Remote Sensing Letters, (SCIE, IF:4.8), 2022 [Paper]

Beat the winner @ CVPRW DeepGlobe Challenge
Introduced a self-attention based operation into a segmentation task.

[J2] Memory-free Stochastic Weight Averaging by One-way Variational Pruning

Yooseung Wang, H. Park, J. Lee

IEEE Signal Processing Letters (SCIE, IF:3.9, Acceptance ≈ 20% ), 2021 [Paper]

[J1] One-shot Summary Prototypical Network Toward Accurate Unpaved Road Semantic Segmentation

Yooseung Wang, D. Lee, J. Heo, J. Park

IEEE Signal Processing Letters (SCIE, IF:3.9, Acceptance ≈ 20% ), 2021 [Paper]

[C2] Transitional Asymmetric Non-local Neural Network for Real-World Dirt Road Segmentation

Yooseung Wang, J. Park

ICPR 2020 [Paper] [Poster] [Slides] [Video]

[C1] Occlusion Data Augmentation for Object detectors based on Random Erasing and Bounding Box Tailoring

J. Heo, Yooseung Wang, J. Park

ICCAS 2020 [Paper]

DOMESTIC PUBLICATIONS

[D6] 왕유승, 우상민, 김창익, "엣지 장치를 위한 경량화된 한국어 음성인식 모델 분석", 대한전자공학회 추계학술대회, 2023
[D5] 이동혁, 왕유승, 허지성, 조선영, 유현성, 박지훈, "전차 부위별 인식 정확도 분석", 한국군사과학기술학회 종합학술대회, 2022
[D4] 왕유승, 조영걸, "은닉 물체 탐지를 위한 자가주의 탐색인지 네트워크", 제 17회 한국로봇종합학술대회 (KRoC), 2022
[D3] 왕유승, 박지훈, "콘볼루션 인공신경망 기반 군용 기동장비 은닉 물체 탐지 기술 연구", 제 31회 영상처리 및 이해에 관한 워크샵 (IPIU), 2022
[D2] 김민수, 김석종, 이기성, 왕유승, "군의 정비능력 향상을 위한 딥러닝 기반의 고장예측체계 활용 연구", 한국군사과학기술학회 종합학술대회, 2021
[D1] 허지성, 이동혁, 왕유승, 임형우, "시뮬레이터를 이용한 자율주행 강화학습 방법 연구", 한국군사과학기술학회 종합학술대회, 2021

PATENT

[PT1] Jihun Park, Hyeonsung Yoo, Yooseung Wang, 의사-레이블 데이터 생성 방법 및 장치, 의사-레이블 데이터를 이용한 심층 신경망 학습 방법 및 장치, "Method and appratus for identifying a driving area of a vehicle using an image and a lidar." (Korea - Registration No. 102336500 / Application No. 10-2021-0015700)
[PT2,] Jihun Park, Hyeonsung Yoo, Yooseung Wang, 영상 및 라이다를 활용한 차량의 주행 가능 영역 식별 장치 및 방법 "Cyclic learning method using heterogeneous sensors and heterogeneous identification algorithms", (Korea - Registration No. 10-2306083 / Application No. 10-2020-0000439)
[PT3] Byeong-Su Kim, Ha-Lin Choi, Yooseung Wang, Sang-Yun Lee, Hee-Seok Kang, “Entry-level 3D Scanner Kit” (Korea - Application No. 10-2017-0179216)

Awards & Honors
2024, 1st Presidential Science Scholarship for Graduate School

2023, Kwanjeong Scholarship (Sole Recipient from KAIST EE/CS/AI)

2022, Kwon Yeong-Se Scholarship (Recipient of Top Academic Entrance Scholarship (1st/2nd Place) in KAIST EE)
2022, 3rd Place, SpaceNet 8 Challenge on CVPRW

2021, 1st Place, AI Idea Competition for Brain Hemorrhage Diagnosis Track

2018, Top 1%, NAVER CEO Award, Han-ium ICT Competition

2018-2019, National Science & Technology Scholarship

2017, UNIST Overseas Studies Scholarship, (5,000 USD)

2016, (Top 2%) 13th place, ACM ICPC Daejeon Regional

2015 4th Place, Global Startup Competition

2015 (Top 2%) 57th out of 2300, LG Code Challenger
2015-2019 UNIST Academic Performance Scholarship (Full Scholarship)

2015 Honorable mention, Korea Super-computing Challenge

Project

2023.02 - 2024.02, Sketch-based Video Object Detection, 스케치 기반 동영상 내 객체 인식 연구
2023.02 - 2024.06, Agency for Defense Development (ADD), Group Activity Recognition, 그룹 행동 인식 연구
2022.08 - 2023.06 Present, Samsung Electronics, Samsung Advanced Institute of Technology (SAIT), 나노 광학 구조 이미지 센서 해상도 개선 알고리즘 연구
2022.01-2022.05, Agency for Defense Development (ADD), Face Recognition

Related Works : IEEE ACCESS (AMM-Softmax)

2019.06-2021.09, Agency for Defense Development (ADD), AI-based Military Autonomous Driving Car for Unpaved Road, 무인체계용 지능형 학습/추론엔진기술

Related Works : ICPR20 (TANNet), T-ITS’21, SPL21 (SWA-OVP), SPL21 (OSPNet)

2019.06-2022.05, Agency for Defense Development (ADD), Occluded Military Vehicle Detection, 인공지능 기반 시각적 부분 가림 물체 자동 식별 기술

Related Works : PRL22 (OSAT), ICCAS20 (ODA)

2018.07-2018.08, Electronics and Telecommunications Research Institute (ETRI), Development of High Performance Visual Discovery Platform for Realtime and Large-Scale Data Analysis and Prediction, 실시간 대규모 영상 데이터 이해·예측을 위한 고성능 비주얼 디스커버리 플랫폼 개발
2018.04-2018.06, UNIST, A machine learning and statistical inference framework for explainable artificial intelligence, 의사결정 이유를 설명할 수 있는 인간 수준의 학습·추론 프레임워크 개발

Related Works : PRL22(SM-VCENet)

Page updated

Google Sites

Report abuse