Research
I am interested in generative models, object-centric learning, and compositional modeling, with focus on generalization to novel settings beyond training data. The applications include visual content generation, scene understanding, physical reasoning, and robotic manipulation.
|
News
- 2025.05 Paper "Compositional Scene Understanding through Inverse Generative Modeling" accepted by ICML 2025.
- 2025.01 I will be joining Qualcomm AI Research Amsterdam as a Research Scientist Intern in the summer of 2025.
- 2024.05 I give a talk on "Compositional Generative Models" at ASML, the Netherlands.
- 2024.04 Paper "Compositional Image Decomposition with Diffusion Models" accepted by ICML 2024.
|
Selected Publications (* stands for equal contribution)
|
|
Compositional Scene Understanding through Inverse Generative Modeling
Yanbo Wang,
Justin Dauwels,
Yilun Du
ICML, 2025
Website /
Paper /
Code
|
|
Compositional Image Decomposition with Diffusion Models
Jocelin Su*,
Nan Liu*,
Yanbo Wang*,
Joshua B. Tenenbaum,
Yilun Du
ICML, 2024
Website /
Paper /
Code
|
|
Slot-VAE: Object-Centric Compositional Image Generation with Slot Attention
Yanbo Wang,
Letao Liu,
Justin Dauwels,
ICML, 2023
Website /
Paper /
Code
|
Research Talks
|
Compositional Inference for Generalizing beyond Training Distribution, SPS Seminar, TU Delft, 2025.
Composing EBMs and Diffusion Models, ASML, Eindhoven, 2024.
Object-Centric Image Generation with Hierarchical VAE, SPS Seminar, TU Delft, 2024
|
Academic Service
|
Reviewer, ICLR 2025
Reviewer, ICML 2025
Reviewer, NeurIPS 2025
|
Teaching Experience
|
Co-instructor, EE4685 Machine Learning, a Bayesian Perspective (Graduate), 2023-2024
Co-instructor, EE4C12 Machine Learning for Electrical Engineering (Graduate), 2022-2024
|
|