| Speaker: | Yin Li (CCA, Flatiron Institute) and Yueying Ni (CMU) |
|---|---|
| Title: | Super-resolution cosmological simulations with deep learning |
| Date (JST): | Thu, May 13, 2021, 11:00 - 12:00 |
| Place: | Zoom |
| Related File: | 2656.pdf |
| Abstract: |
Cosmological simulations are indispensable for understanding our Universe. The large dynamic range incurs large computational costs, demanding sacrifice of either resolution or size, and often both. We build a deep neural network to enhance low-resolution dark matter simulations, generating super-resolution realizations that agree remarkably well with authentic high-resolution counterparts on various statistics, and are orders-of-magnitude faster. It readily applies to larger volumes, and generalizes to rare objects not present in the training data. We show that deep learning and cosmological simulations can be a powerful combination, to model the structure formation of our Universe over its full dynamic range. The talk is based on our recent works: https://arxiv.org/abs/2010.06608 https://arxiv.org/abs/2105.01016 |
