Segment clouds in satellite infrared single channel images.
Supervised training of a model to learn the segmentation mask as a target labels (Y) from the satellite images (X).
Pairs of IR (single channel) images and binary masks for supervised learning.
- Image/mask size: 1024x1024x1
- Approx. 1000 samples
- TIFF image format
- Adaption to multi channel base model, by replicating gray scale input image to 3 channels
- Augmentations (rotation, flip, noise, cropping)
Transfer learning for image semantic segmentation tasks
Location: /tensorflow_vgg/
(static preprocessed TFRecord dataset, no augmentation)
- Flat (one step) upsampling decoder
- Deeper decoder designs increase training challenge significantly
Location: /pytorch_segformer/
- Two step upsampling decoder