Published June 22, 2023
| Version v1.0
Software
Open
jaschers/psnet: v1.0
Authors/Creators
- 1. University of Groningen
Description
PSNet is the first application of pattern spectra on convolutional neural networks (CNNs) for the event reconstruction of imaging atmospheric Cherenkov telescopes (IACTs). We train a CNN on pattern spectra of gamma-ray events from the Cherenkov Telescope Array (CTA) for energy reconstruction and signal-background separation. PSNet is based on Tensorflow 2.3.1 and Keras 2.4.3 and uses the ctapipe software for the data handling.
Files
jaschers/psnet-v1.0.zip
Files
(9.7 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:af7fd29520f807a98eeb3fa125052ff5
|
9.7 MB | Preview Download |
Additional details
Related works
- Is supplement to
- https://github.com/jaschers/psnet/tree/v1.0 (URL)
References
- J. Aschersleben et al. (2021). arxiv.org/abs/2108.00834
- J. Aschersleben et al. (2023). arxiv.org/abs/2302.11876