15th Prediction Science Seminar
- Date
- April 6th, 2023 (Thu.) 10:00-11:30 (JST)
- Language
- Engish
- Place
- Zoom
To join the seminar, please contact the Prediction Science Seminar Office: prediction-seminar[remove here]@ml.riken.jp
Program
| Time | Content | Speaker |
|---|---|---|
| 10:00-11:00 | Recent trends in deep generative models and world models | Dr. Masahiro Suzuki (University of Tokyo) |
| 11:00-11:30 | Discussion | - |
Abstract
Deep generative models are a powerful framework for modeling the process of data generation using deep neural networks. These models have been applied to a wide range of areas, including image and document generation, representation learning, anomaly detection, and more. In this talk, I will provide an overview of recent research on deep generative models, and present our work on multimodal learning using deep generative models and developing a library of deep generative models. I will also introduce world models as an extension of deep generative models and discuss recent research trends in these areas, including our recent work on object-centric world models.
Organizer
- Prediction Science Laboratory (RIKEN CPR)
Co-organizer
- Data Assimilation Research Team (R-CCS)
- RIKEN Interdisciplinary Theoretical and Mathematical Sciences Program
- Environmental Metabolic Analysis Research Team (RIKEN CSRS)
- Computational Climate Science Research Team (R-CCS)
- Medical Data Deep Learning Team (R-IH)
- Medical Data Mathematical Reasoning Team (R-IH)
- Laboratory for Physical Biology (RIKEN BDR)