eduzhai > Applied Sciences > Engineering >

Image Colorization A Survey and Dataset

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 19 pages

Abstract: Image colorization is an essential image processing and computer visionbranch to colorize images and videos. Recently, deep learning techniquesprogressed notably for image colorization. This article presents acomprehensive survey of recent state-of-the-art colorization using deeplearning algorithms, describing their fundamental block architectures in termsof skip connections, input etc. as well as optimizers, loss functions, trainingprotocols, and training data etc. Generally, we can roughly categorize theexisting colorization techniques into seven classes. Besides, we also providesome additional essential issues, such as benchmark datasets and evaluationmetrics. We also introduce a new dataset specific to colorization and performan experimental evaluation of the publicly available methods. In the lastsection, we discuss the limitations, possible solutions, and future researchdirections of the rapidly evolving topic of deep image colorization that thecommunity should further address. Dataset and Codes for evaluation will bepublicly available at this https URL

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×