Scholar
Sanghyun Woo
Google Scholar ID: iwBPvPIAAAAJ
Google DeepMind
Citations & Impact
Citations
38,188
H-index
24
i10-index
35
Publications
20
Co-authors
25
list available
Publications
Resume (English only)
Academic Achievements
- Publications:
- - BlenderFusion: 3D-Grounded Visual Editing and Generative Compositing
- - ε-VAE: Denoising as Visual Decoding
- - Cambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMs
- - SwitchLight: Co-design of Physics-driven Architecture and Pre-training Framework for Human Portrait Relighting
- - MTMMC: A Large-Scale Real-World Multi-Modal Camera Tracking Benchmark
- - Test-time Adaptation in the Dynamic World with Compound Domain Knowledge Management
- - ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders
- - Mask-guided Matting in the Wild
- - Bidirectional Domain Mixup for Domain Adaptive Semantic Segmentation
- - Learning Classifiers of Prototypes and Reciprocal Points for Universal Domain Adaptation
- - Bridging Images and Videos: A Simple Learning Framework for Large Vocabulary Video Object Detection
- Awards: Microsoft Research Asia PhD Fellowship, Google Research PhD Fellowship
Research Experience
- Interned at Adobe Research (San Jose, CA) and Meta AI (Menlo Park, CA). Previously, a Faculty Fellow in Computer Science at NYU Courant, hosted by Prof. Saining Xie.
Education
- Ph.D. and M.S. degrees in electrical engineering from Korea Advanced Institute of Science and Technology (KAIST), advised by Prof. In So Kweon; B.S. degree in electrical computer engineering from Seoul National University (SNU).
Background
- Research interests: designing effective visual models to better understand the world and developing efficient learning frameworks to utilize data at scale with minimal human supervision. Currently a Senior Research Scientist at Google DeepMind.
