SUnCNN: Sparse Unmixing Using Unsupervised Convolutional Neural Network SUnCNN is the first deep learning-based technique proposed for sparse unmixing. It uses a deep convolutional encoder-decoder to generate the abundances relying on a spectral library. We reformulate the sparse unmixing into an optimization over the deep network’s parameters. Therefore, the deep network learns in an unsupervised manner to map a fixed input intothe sparse optimum abundances. Additionally, SUnCNN holds the sum-to-one constraint using a softmax activation layer.
Note that the model used is a modified version of the DIP software (https://github.com/DmitryUlyanov/deep-image-prior) which is uploaded here and therefore the copyright of it is preseved.
Note that the results reported in the paper are mean values over ten experiments. tol2 variable is the number of runs.
If you use this code please cite the following paper Rasti, B., and Koirala, "SUnCNN: Sparse Unmixing Using Unsupervised Convolutional Neural Network" IEEE Geoscience and Remote Sensing Letters.
DC1: The fractional abundance of endmember 2. From top to bottom SNR of 20, 30, and 40 dB.
