Ecosyste.ms: Repos
An open API service providing repository metadata for many open source software ecosystems.
GitHub / vedantk-b / Cloud-Segmentation-from-Satellite-Imagery
This repository contains the first model which I tried on cloud dataset from sentinel satellite for cloud segmentation. The model was UNET and the loss function used was BinaryCrossEntropy with Logits. For, validation IoU was used. Accuracy of 81.39% could be achieved on the validation set using this model.
Stars: 7
Forks: 4
Open Issues: 0
License: None
Language: Python
Repo Size: 625 KB
Dependencies:
13
Created: over 2 years ago
Updated: about 1 year ago
Last pushed: about 2 years ago
Last synced: 8 months ago
Files
Dependencies
- cloudpathlib ==0.6.2
- crfseg ==1.0.0
- lightning-flash ==0.5.0
- mmcv-full ==1.3.9
- mmsegmentation ==0.16.0
- nnunet ==1.7.0
- pandas_path ==0.3.0
- planetary-computer ==0.4.2
- rioxarray ==0.9.0
- segmentation-models ==1.0.1
- tensorflow-gpu ==2.6.0
- ttach ==0.0.3
- xarray-spatial ==0.3.0