About Me

I am Dong-Ha Shin, a first-year M.S. student in the Computational Imaging Group, led by Prof. Seung-Hwan Baek.

During my undergraduate studies, I was affiliated with the Integrated Plasmonics and Optical Device Laboratory, advised by Prof. Seung-Yeol Lee.

My research interests broadly lie in co-designing novel systems at the intersection of optics, vision, graphics, and artificial intelligence. I believe the multi-dimensional nature of light holds expressive power and untapped potential that extend well beyond conventional 2D intensity images.

As part of my military service, I led the development of a globally deployed Road-Information-as-a-Service (RiaaS)—currently operating in real-world environments across cities including New York City and Seoul— encompassing edge-device systems, AI model training, on-device AI deployment, and full web services. Additionally, through the NamuhX project, I collaborated with Qualcomm, designing and implementing both an on-device NPU acceleration pipeline and a voice-interactive AI agent.

Dong-Ha Shin

Publications

PADO: PyTorch Automatic Differentiable Optics
Open-sourceLead developer and main maintainer

PADO: PyTorch Automatic Differentiable Optics

POSTECH Computational Imaging Group

2025

Complex-Valued Holographic Radiance Fields
Journals

Complex-Valued Holographic Radiance Fields

Yicheng Zhan, Dong-Ha Shin, Seung-Hwan Baek, and Kaan Akşit

arXiv 2025 (under review)

Hologram Upscaling for Viewing Angle Expansion Using Light Field Extrapolation with Object Detection Algorithm
Journals

Hologram Upscaling for Viewing Angle Expansion Using Light Field Extrapolation with Object Detection Algorithm

Dong-Ha Shin, Chee-Hyeok Song, and Seung-Yeol Lee

Current Optics and Photonics, Vol. 9, Issue 1, pp. 55–64, 2025

A Technique for Interpreting and Adjusting Depth Information of each Plane by Applying an Object Detection Algorithm to Multi-plane Light-field Image Converted from Hologram Image
Journals

A Technique for Interpreting and Adjusting Depth Information of each Plane by Applying an Object Detection Algorithm to Multi-plane Light-field Image Converted from Hologram Image

Young-Gyu Bae, Dong-Ha Shin, and Seung-Yeol Lee

Journal of Broadcast Engineering, Vol. 28, No. 1, January 2023