I-Hsiang Chen
Aaron
F09921058@g.ntu.edu.tw
+1 206 229 8445
SKILLS
Research
Unified Multimodal Model
Image Generation
Robust Learning
Information Retrieval
Programming
Python
C
Verilog
MATLAB
PCB Layout
Soft Power
Team Leader
Presentation
Notion Organization
InDesign
Certificate
TQC+ Python
Instrument Electronic Class B
Industrial Electronics Class C
SCHOLARSHIPS
2025 NTU Outstanding Young Award
2023 CTCI Research Award
2022 Hon-Hai Tech Award
2021 Advanced Technology Award
2020 ASML Elite Scholarship
2020 Hon-Hai Whale Scholarship
EDUCATION
University of Washington | Visiting Scholar May. 2024 - Present
Electrical and Computer Engineering;; Topic: Diffusion Model, Image enhancement; Advisor: enq-Neng Hwang
National Taiwan University | Ph.D. Sep. 2020 - Present
Electrical Engineering and Computer Science; Topic: Image enhancement, Re-identification, Crowd Counting; Advisor: Sy-Yen Kuo, GPA: 4.23/4.3
National Taiwan Tech | Bachelor Sep. 2016 - Jun. 2020
Electronic and Computer Engineering; minor: Bachelor Degree Program of Finance, concentration in IC Design, MCU, GPA: 4.09/4.3
PUBLICATION
Domain Generalization "Exploring probabilistic modeling beyond domain generalization for semantic segmentation", In ICCV, 2025. (First author)
Image Restoration "UniRestore: Unified Perceptual and Task-Oriented Image Restoration Model", In CVPR Highlight, 2025. (First author)
Crowd Counting "Improving Point-based Crowd Counting and Localization Based on Auxiliary Point Guidance", In ECCV, 2024. (First author)
Haze Removal "Semantic Guidance Learning for High-Resolution Non-Homogeneous Dehazing", In CVPRW, 2023. (Second author)
Shadow Removal "TSRFormer: Transformer based two-stage refinement for single image shadow removal", In CVPRW, 2023. (Second author)
Re-identification "RVSL: Robust Vehicle Similarity Learning in Real Hazy Scenes Based on Semi-supervised Learning", In ECCV, 2022. (First author*)
Re-identification "SJDL-Vehicle: Semi-supervised Joint Defogging Learning for Foggy Vehicle Re-identification", In AAAI, 2022. (First author*)
Reflection Removal "Missing Recovery: Single Image Reflection Removal based on Auxiliary Prior Learning", IEEE TIP, 2022.(Second author)
Snow Removal "ALL Snow Removed: Single Image Desnowing Algorithm Using Hierarchical Dual-tree Complex Wavelet Representation and Contradict Channel Loss", In ICCV, 2021. (Second author)
Rain Removal "ContourletNet: A Generalized Rain Removal Architecture Using Multi-Direction Representation and Hierarchical Decomposition", In BMVC, 2021. (Second author)
Smoke Removal "DesmokeNet: A Two-stage Smoke Removal Pipeline Based on Self-Attentive Feature Consensus and Multi-Level Contrastive Regularization", IEEE TCVST, 2021. (Second author)
EXPERIENCE
Samsung Research UK | Research Intern Jun. – Dec. 2025
Conducted research on multimodal understanding and generation model.
MTK MM | Research Intern Jul. – Dec. 2023
Research and Development of Video Codec with Deep Learning
ASML D&E | Research Intern Jan. – Apr. 2022
Develop a negative image generation model for improved precision of eScan
Capacura Inc. (Cologne, Germany)| Intern Jul. – Sep. 2019
Develop an IPO venture capital forecast by categories and trends features
Industrial Technology Research Institute | Contractor Dec. 2018 – Sep. 2019
Assist data mining and organization for self-driving development
SIEMENS (Berlin, Germany) | Workshop Aug. 2016
Summer workshop for programmable logic controller training
COMPETITION
ADVANTECH AIoT InnoWorks | 3rd Place Feb. – Sep. 2019
Develop a campus restaurant management system with object detection, microcontroller, and monitor
TQC+ Python Programming Competition | Merit Award May. 2019
Large-scale competition of programming ability of colleges and universities
Senior High School Project Contest | Honorable Mention Fall 2016
Develop a machine-learning bicycle training system with microcontrollers
WorldSkills Competition (Taiwan) | 1st Place Aug. 2014 – Sep. 2015
Circuit design, Microcontroller programming, and Function design