Super-Resolution Generative Adversarial Network for Data Compression of Direct Numerical Simulations
Guardat en:
| Publicat a: | arXiv.org (Dec 18, 2024), p. n/a |
|---|---|
| Autor principal: | |
| Altres autors: | , , , , , , |
| Publicat: |
Cornell University Library, arXiv.org
|
| Matèries: | |
| Accés en línia: | Citation/Abstract Full text outside of ProQuest |
| Etiquetes: |
Sense etiquetes, Sigues el primer a etiquetar aquest registre!
|
MARC
| LEADER | 00000nab a2200000uu 4500 | ||
|---|---|---|---|
| 001 | 3147267066 | ||
| 003 | UK-CbPIL | ||
| 022 | |a 2331-8422 | ||
| 035 | |a 3147267066 | ||
| 045 | 0 | |b d20241218 | |
| 100 | 1 | |a Nista, Ludovico | |
| 245 | 1 | |a Super-Resolution Generative Adversarial Network for Data Compression of Direct Numerical Simulations | |
| 260 | |b Cornell University Library, arXiv.org |c Dec 18, 2024 | ||
| 513 | |a Working Paper | ||
| 520 | 3 | |a The advancement of high-performance computing has enabled the generation of large direct numerical simulation (DNS) datasets of turbulent flows, driving the need for efficient compression/decompression techniques that reduce storage demands while maintaining fidelity. Traditional methods, such as the discrete wavelet transform, cannot achieve compression ratios of 8 or higher for complex turbulent flows without introducing significant encoding/decoding errors. On the other hand, super-resolution generative adversarial networks (SR-GANs) can accurately reconstruct fine-scale features, preserving velocity gradients and structural details, even at a compression ratio of 512, thanks to the adversarial training enabled by the discriminator. Their high training time is significantly reduced with a progressive transfer learning approach and, once trained, they can be applied independently of the Reynolds number. It is demonstrated that SR-GANs can enhance dataset temporal resolution without additional simulation overhead by generating high-quality intermediate fields from compressed snapshots. The SR-GAN discriminator can reliably evaluate the quality of decoded fields, ensuring fidelity even in the absence of original DNS fields. Hence, SR-GAN-based compression/decompression methods present a highly efficient and scalable alternative for large-scale DNS storage and transfer, offering substantial advantages in terms of compression efficiency, reconstruction fidelity, and temporal resolution enhancement. | |
| 653 | |a Datasets | ||
| 653 | |a Temporal resolution | ||
| 653 | |a Wavelet transforms | ||
| 653 | |a Velocity gradient | ||
| 653 | |a Data compression | ||
| 653 | |a Discriminators | ||
| 653 | |a Discrete Wavelet Transform | ||
| 653 | |a Generative adversarial networks | ||
| 653 | |a Fluid flow | ||
| 653 | |a Domain names | ||
| 653 | |a Encoding-Decoding | ||
| 653 | |a Direct numerical simulation | ||
| 653 | |a Reynolds number | ||
| 653 | |a Fluid dynamics | ||
| 653 | |a Compression ratio | ||
| 700 | 1 | |a Schumann, Christoph D K | |
| 700 | 1 | |a Fröde, Fabian | |
| 700 | 1 | |a Gowely, Mohamed | |
| 700 | 1 | |a Grenga, Temistocle | |
| 700 | 1 | |a MacArt, Jonathan F | |
| 700 | 1 | |a Attili, Antonio | |
| 700 | 1 | |a Pitsch, Heinz | |
| 773 | 0 | |t arXiv.org |g (Dec 18, 2024), p. n/a | |
| 786 | 0 | |d ProQuest |t Engineering Database | |
| 856 | 4 | 1 | |3 Citation/Abstract |u https://www.proquest.com/docview/3147267066/abstract/embedded/6A8EOT78XXH2IG52?source=fedsrch |
| 856 | 4 | 0 | |3 Full text outside of ProQuest |u http://arxiv.org/abs/2412.14150 |