Patsorn Sangkloy
researchpublicationscvcontact

Patsorn Sangkloy.

Researcher · Computer vision & machine learning · SF

Patsorn is a researcher in computer vision and machine learning with a decade of experience across academia and industry. She is currently obsessed with building a more semantically structured world model at Eigen 4D.

Google Scholar CV (PDF) patsorn.sangkloy [at] gmail.com
Patsorn Sangkloy
Earlier
work

Over the previous decade, Patsorn's research helped make generative vision controllable and grounded — from some of the earliest deep networks for user-guided image synthesis, to datasets and benchmarks now standard across the field, to models of how people attend to and judge generated content. Carried out with Adobe Research, Argo AI, Georgia Tech, and NYU, this work has yielded a U.S. patent, oral and spotlight presentations at top venues, and coverage from TechCrunch to New Scientist.

Selected
publications
Infinite Gaze Generation for Videos with Autoregressive Diffusion. ECCV 2026.
Cost-Aware Routing for Efficient Text-to-Image Generation. TMLR 2026.
GeneVA: Human Annotations for Generative Text-to-Video Artifacts. WACV 2026.
StyleGAN Salon: Multi-View Latent Optimization for Pose-Invariant Hairstyle Transfer. CVPR 2023.
A Sketch Is Worth a Thousand Words: Image Retrieval with Text and Sketch. ECCV 2022.
Argoverse: 3D Tracking and Forecasting with Rich Maps. CVPR 2019, oral.
Kernel Mean Matching for Content Addressability of GANs. ICML 2019, oral.
TextureGAN: Controlling Deep Image Synthesis with Texture Patches. CVPR 2018, spotlight.
Scribbler: Controlling Deep Image Synthesis with Sketch and Color. CVPR 2017.
The Sketchy Database: Learning to Retrieve Badly Drawn Bunnies. SIGGRAPH 2016.
Complete list on Google Scholar →
Education
PhD, Computer Science — Georgia Institute of Technology · advisor: James Hays
ScM, Computer Science — Brown University
BEng, Computer Engineering — Khon Kaen University · first-class honors and gold medal, first of 649
Recognition
Adobe Research Fellowship2018
U.S. Patent 10,916,001 — sketch-conditioned image synthesis; presented at Adobe MAX2017–21
Oral presentations at CVPR and ICML; spotlight at CVPR 20182018–19
Royal Thai Government Scholarship2013–19
Honor Medal, conferred by HRH the Crown Prince of Thailand; graduated first of 6492012
Co-organizer, Women in Computer Vision (ICCV 2023, ECCV 2024); Machine Learning Research School, Bangkok2023–24
Contact

Correspondence and research collaboration:

patsorn.sangkloy [at] gmail.com

© 2026 Patsorn Sangkloy