Image inpainting aims to complete the missing or corrupted regions of images with realistic contents . WaveFill decomposes images by using discrete wavelettransform (DWT) that preserves spatial information naturally . It applies L1reconstruction loss to the decomposed low-frequency bands and adversarial loss . It effectively mitigate inter-frequency conflicts while completing images in spatial domain . Extensive experiments over multipledatasets show that WaveFill achieves superior image inpaining qualitatively and quantitatively . It is a novel normalization scheme that aligns and fuses themulti-frequency features effectively, we design a novel Normalization scheme . It uses a wavelet-based network that fuses features with distinctstatistics and aligns themultulti

Author(s) : Yingchen Yu, Fangneng Zhan, Shijian Lu, Jianxiong Pan, Feiying Ma, Xuansong Xie, Chunyan Miao

Links : PDF - Abstract

Code :
Coursera

Keywords : wavefill - image - frequency - images - effectively -

Leave a Reply

Your email address will not be published. Required fields are marked *