Internal Diverse Image Completion

Image completion is widely used in photo restoration and editing applications, e.g. for object removal. Recently, there has been a surge of research on generating diverse completions for missing regions. However, existing approaches require large training sets from a specific domain of interest, and often fail to provide satisfactory results for general-content images. In this talk I will present a diverse completion method that does not require any training set and can thus treat arbitrary images from any domain. Our internal diverse completion (IDC) approach draws inspiration from recent single-image generative models that are trained on multiple scales of a single image, adapting them to the extreme setting where only a small portion of the image is available for training. We illustrate the strength of IDC on several datasets, using both user studies and quantitative comparisons.

* M.Sc. student under the supervision of Professor Tomer Michaeli.