With the development of the deep learning or intelligent algorithms, many traditional problems in image processing has been redesigned from the deep learning view and achieved good performance, such as image restoration, single image super-resolution, image in-painting, and image coding. The purpose of this workshop is to bring together the researchers in image processing and intelligent algorithms to discuss recent improvements in these areas.
In the image super-resolution area, the aim is to use a low-resolution image to produce a high-resolution image. Thus, it can increase the clarity and magnify an image. Since the SRCNN algorithm, which uses convolution neural network, many intelligent algorithms have been proposed for the image super-resolution and they have achieved better performance than traditional algorithms, especially for the human face image super-resolution. They also have many application areas in machine visions, which is a major topic of the conference.
In the image in-painting area, the aim is to use an image with missing information or a hole in the image to estimate the missing information or the pixel values inside the image hole. Many intelligent algorithms have been proposed in this area which use an encoder and decoder by deep learning. Any intelligent algorithm in the image or video processing area is welcome in the workshop.