Navigation Structure
Table of contents
Tensorflow Test Time Augmentation (tensorflow-tta)
Tensorflow Test Time Augmentation: Test time augmentation on Tensorflow keras models which utilizes GPU acceleration for segmentation models. It helps in improving accuracy of the model over the test dataset
Change Detection with Google Earth Engine
This repo contains filtering modules for dynamic world data available on google earth engine. The filtering functions have been written to aid in time-series analysis and solve problems such as poor predictions due to cloudy or unforseen weather conditions through the utilisation previous trust-worthy values. It can also be used for the Sentinel-2 data catalogue. It also contains example codes on change detection post filtering the data with the modules available in this repo. The data thus generated can be visualized with a flask application that can be run with the main.py file in the app folder.