Research Assistant, CITI, Academia Sinica
I-Sheng Fang
Ethan Fang / 方 宜晟 / Gî-Tshiânn Png
"I-Sheng" is pronounced like "Ethan".
I am currently exploring Fall 2027 PhD opportunities in computer vision, generative AI, and photography-aware visual intelligence.
News
Recent Updates
Adapt2Hide is accepted by ICIP 2026.
The Generative AI for Photography (GAIP) Workshop was successfully held at WACV 2026. Thanks to all speakers, authors, reviewers, and participants!
Gen-n-Val is accepted by CVPR 2026 Findings track. See you in Denver!
I am honored to be selected for Taiwan Government Fellowship for Study Abroad(公費留考), which provides full support in 3 years for my future PhD study abroad.
About
Researcher working where vision models meet creative tools
I am a Research Assistant in CITI at Academia Sinica, working with Dr. Jun-Cheng Chen on generative models and computer vision. I received my Master's degree in Robotics from National Yang Ming Chiao Tung University in January 2023, advised by Prof. Yong-Sheng Chen and Prof. Wei-Chen (Walon) Chiu. Before that, I studied computer science at National Chengchi University with Prof. Yan-Tsung Peng, worked as a research assistant at the Enriched Vision Applications Lab at National Chiao Tung University, and received my Bachelor's degree in Mathematical Science from National Chengchi University in January 2018.
Publications
Selected Publications
Peer-reviewed and workshop work in generative modeling, camera-aware synthesis, depth sensing, stereo matching, and style transfer.
US\(^{3}\)Net: Ultralightweight Self-Supervised Stereo Matching Network using Depth-Aware Geometric Soft Occlusion
An ultralightweight self-supervised stereo matching network for efficient stereo depth estimation on resource-constrained devices, combining a low-complexity feature extractor with Depth-Aware Geometric Soft Occlusion (DAGSO) to improve occlusion handling while using only 12K parameters.
Adapt2Hide: Leveraging Off-the-Shelf Autoencoder for Reversible Visual Processing
* Equal contribution.
An image steganography approach using large pre-trained autoencoders and LoRA to support high-quality message reconstruction with minimal additional parameters for reversible visual processing applications.
Gen-n-Val: Agentic Image Data Generation and Validation
* Equal contribution.
An agentic framework for image data generation and validation using Layer Diffusion, LLM prompt agents, and VLLM validation agents to improve object detection and segmentation training data.
ImagenWorld: Stress-Testing Image Generation Models with Explainable Human Evaluation on Open-ended Real-World Tasks
* Equal contribution.
A large-scale benchmark for evaluating modern image-generation models across open-ended real-world tasks with fine-grained human annotations.
Every Camera Effect, Every Time, All at Once: 4D Gaussian Ray Tracing for Physics-based Camera Effect Data Generation
* Equal contribution. Work done at Academia Sinica as interns. † Internship mentor.
A two-stage pipeline combining 4D Gaussian Splatting with physically based ray tracing to simulate real-world camera effects such as fisheye distortion, rolling shutter, and depth of field.
Single Image Reflection Removal Based on Knowledge-Distilling Content Disentanglement
A single-image reflection removal method that disentangles reflection and transmission features with knowledge distillation.
Projects
Creative, Research, and Open Source Work
A Century of Heartfelt Sentiment: 100th Anniversary Special Exhibition of the Taiwan Cultural Association
Exhibition Room D, National Museum of Taiwan Literature, Tainan, Taiwan
Deepfake video synthesis
Synthesized DeepFake video for historical figures of the Taiwan Cultural Association as part of a museum exhibition.
Typography Research Collection
Curator
A research collection at the intersection of typography, computer graphics, computer vision, and machine learning.
Awesome Generative AI for Photography
Curator
A curated list of research that integrates photographic principles, concepts, techniques, and domain knowledge into image and video generative models.
Experience
Experience Snapshot
March 2024 - Present
Research Assistant
Research Center for Information Technology Innovation (CITI), Academia Sinica
Working on generative models with Dr. Jun-Cheng Chen.
March 2022 - November 2022
Software Engineer Intern
Microsoft AI R&D Center
Worked on vision transformers, perceptual loss, and generative models with SunDa Yang, Chien-Yi Wang, Prof. Shang-Hong Lai, Dr. Trista Chen, and the face science team.
September 2018 - September 2019
Research Assistant
Enriched Vision Applications Lab, National Chiao Tung University
Worked on style transfer and generative models with Prof. Wei-Chen (Walon) Chiu.
Service
Reviewing and Organizing
Reviewer / 2024 - 2026
Conference Reviewing
NeurIPS 2024, CVPR 2025, ICCV 2025, NeurIPS 2025, WACV 2026, ICRA 2026, CVPR 2026, ICML 2026, ECCV 2026, SIGGRAPH Asia 2026.
Organizer / 2026
Workshop on Generative AI for Photography, WACV 2026
Primary and contact organizer.
Administrator / Ongoing
Enjoyfonts
Administrator for a typography-focused Facebook group.
Personal
Outside the Lab
My work is technical, but my eye is shaped by cameras, letterforms, airports, ballparks, and gyms.
@ishengfang
Instagram
Contact