Pakkapon Phongthawee
Computer Vision & HPC Systems
I am a Ph.D. candidate in the Vision and Learning Lab (VLL) at VISTEC in Rayong, Thailand, advised by Prof. Supasorn Suwajanakorn. My research focuses on advancing state-of-the-art neural rendering, lighting estimation, and diffusion models. Concurrently, I architect and manage the primary HPC infrastructure for both our lab and the wider institute, overseeing a 200+ GPU compute cluster that powers large-scale AI research
Publications & Research
DiffusionLight-Turbo TPAMI 2026
Accelerates the estimation of high-dynamic-range (HDR) environment maps to deliver accurate scene lighting in under 30 seconds.
DiffusionLight CVPR 2024 (ORAL)
Estimates realistic scene illumination by virtually in-painting a reflective chrome ball into any standard image.
StyleGAN Salon CVPR 2023
Performs high-fidelity, structure-aware hairstyle transfer seamlessly adapting reference hair onto target portrait images.
NeX360 TPAMI 2022
Achieves real-time, 360-degree novel view synthesis for highly detailed object-centric scenes.
NeX CVPR 2021 (ORAL)
Enables real-time novel view synthesis of complex forward-facing scenes utilizing neural basis expansion.
HPC & AI Compute Management
- Server Admin: Deployment, hardware monitoring, and thermal management.
- Orchestration: Slurm, Docker/Singularity environments with CUDA GPUs.
- Storage: High-throughput dataset I/O optimization using BeeGFS and NFS.
- Data Distribution: Architected and maintain a geo-distributed dual-mirror file system (east.ph.in.th / west.ph.in.th) to reliably serve large-scale datasets for sharing my research.
- Web: Maintaining and coding web applications using LAMP stack and modern JavaScript frameworks.
Skills & Tools
AI & Computer Vision
HPC & Infrastructure
Experience & Education
Experience
2019 - Present
HPC Admin @ VISTEC
Education
2019 - Present
Ph.D. in Information Science and Technology @ VISTEC
2015 - 2019
B.Sc in Mathematics @ Silpakorn University
GPA: 3.62 / 4.00 (First Class Honors)