(1) Deep learning theory and its applications. Researching the structure design of convolutional neural network, optimization algorithm, and neural architecture search, etc. Furthermore, we apply these techniques to image contour detection, semantic segmentation, image classification, etc. (2) Image segmentation theory and its application. Studying feature extraction, feature dimension-reduction, supervision learning, unsupervision learning, transfer learning, etc. Using these techniques to achieve image superpixel segmentation, contour detection, interactive objection detection, etc.
(3) Image understanding theory and its applications. Studying model-driven, data-driven, model-driven jointing data-driven theory and method to achieve image semantic segmentation, objection recognition, etc.
(4) Image superresolution. Studying image super-resolution algorithms based on sparse representation, neural network, transfer learning, etc, and applying these algorithms to the fields of computer vision, remote sensing, medicine, etc.
|