https://github.com/andrewannex/unet-keras
U-Net model for Keras
Science Score: 10.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
○codemeta.json file
-
○.zenodo.json file
-
○DOI references
-
✓Academic publication links
Links to: arxiv.org -
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (3.4%) to scientific vocabulary
Last synced: 9 months ago
·
JSON representation
Repository
U-Net model for Keras
Basic Info
Statistics
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
- Releases: 0
Fork of pietz/unet-keras
Created about 8 years ago
· Last pushed almost 9 years ago
https://github.com/AndrewAnnex/unet-keras/blob/master/
# U-Net for Keras This is an implementation of the U-Net model for Keras. When using the default parameters this will be the same as the original architecture, except that this code will use padded convolutions. Also, the original paper does not state the amount of Dropout used. Since 0.5 is a common choice in the industry, it was used as the default value. If you are aware of other differences, please contact me. U-Net: Convolutional Networks for Image Segmentation (https://arxiv.org/abs/1505.04597) Other than that, I added a few bells and whistles. This implementation supports transposed convolutions (deconvolutions) for upsampling, strided convolutions for downsampling, batchnormalization and residual connections. If you have further ideas, let me know.
Owner
- Name: Dr. Andrew Annex
- Login: AndrewAnnex
- Kind: user
- Location: Pasadena, CA
- Company: SETI Institute
- Website: http://andrewannex.com
- Twitter: AndrewAnnex
- Repositories: 179
- Profile: https://github.com/AndrewAnnex
Coder and Planetary Science PhD from Johns Hopkins University. Fmr Postdoc @ Caltech, Current Senior Science Systems Engineer @ SETI Institute