Projects for the course Geospatial Big Data Analysis offered by the Master's program in Data Science and Machine Learning of the National Technical University of Athens.
The contents of the projects are summarized below:
- Analysis on multispectral satellite images with Google Earth Engine.
- Vector and raster data exploration using Open Geospatial Consortium protocols, geopandas and leaflet.
- Image segmentation on multispectral satellite images with pixel-wise and patch-wise approaches.
- Video classification with RNNs and Transformers.
- Object detection on natural images with RetinaNet and Faster RCNN.
- Rotated object detection on aerial images with Rotated RetinaNet and Oriented RCNN, using MMRotate.