Deep learning-based hyperspectral image (HSI) classification methods have recently shown excellent performance, however, there are two shortcomings that need to be addressed. One is that deep network training requires a large number of labeled images, and the other is that deep network needs to learn a large number of parameters. They are also general problems of deep networks, especially in the applications that require professional techniques to acquire and label images, such as HSI and medical images. In this talk, two deep network architecture, SAFDNet and DSAFDNet based on the recently developed mathematical theory – stochastic adaptive Fourier decomposition (SAFD) will be introduced. The difference between the proposed deep networks and the general deep convolution networks is that the convolution kernels of the former are learned through SAFD, while the convolution kernels of the latter are learned through back propagation (BP). Since the convolution kernels obtained by SAFD decomposition are complex numbers, few deep learning methods directly deal with such complex convolution kernels. The difference between SAFDNet and DSAFDNet is that the former uses only the real parts of complex numbers as convolutional kernels, and the later uses both real and imaginary parts of complex numbers as convolutional kernels. SAFD has powerful unsupervised feature extraction capabilities, so the two entire deep networks only require a small number of annotated images to train the classifier. In addition, we use fewer convolution kernels in the entire deep networks, which greatly reduces the number of deep network parameters. Experimental results on three popular HSI classification datasets show that our proposed deep network structures outperform other compared state-of-theart deep learning methods in HSI classification. 

Liming Zhang received the B.S. degree in Computer Software at Nankai University, China and M.S. degree in Signal Processing at Nanjing University of Science and Technology, China. She received her PhD degree in image processing at University of New England, Australia. She is currently an assistant professor in Faculty of Science and Technology, University of Macau. Her research interests include Computer vision, Image processing, Artificial intelligence, Machine learning, and Deep learning. She has published over 100 papers, including IEEE Transactions on Image Processing, IEEE Transactions on Signal Processing, CVPR, ect. The main contribution lies on new image and signal processing methodology – adaptive Fourier decomposition (AFD)based image processing methods and new deep network development. The image and video compression results based on stochastic AFD (SAFD) exceed the current international image and video compression standards JPEG, JPEG2000, MPEG, and also exceed the compression results of the popular deep networks. SAFD-based deep networks also perform well in alleviating the scarcity of training data.