{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/sampling-using-neural-networks-for-colorizing","title":"Sampling Using Neural Networks for colorizing the grayscale images","arxiv_id":"1812.10650","date":"2018-12-27","proceeding":null,"authors":["Wonbong Jang"],"abstract":"The main idea of this paper is to explore the possibilities of generating\nsamples from the neural networks, mostly focusing on the colorization of the\ngrey-scale images. I will compare the existing methods for colorization and\nexplore the possibilities of using new generative modeling to the task of\ncolorization. The contributions of this paper are to compare the existing\nstructures with similar generating structures(Decoders) and to apply the novel\nstructures including Conditional VAE(CVAE), Conditional Wasserstein GAN with\nGradient Penalty(CWGAN-GP), CWGAN-GP with L1 reconstruction loss, Adversarial\nGenerative Encoders(AGE) and Introspective VAE(IVAE). I trained these models\nusing CIFAR-10 images. To measure the performance, I use Inception Score(IS)\nwhich measures how distinctive each image is and how diverse overall samples\nare as well as human eyes for CIFAR-10 images. It turns out that CVAE with L1\nreconstruction loss and IVAE achieve the highest score in IS. CWGAN-GP with L1\ntends to learn faster than CWGAN-GP, but IS does not increase from CWGAN-GP.\nCWGAN-GP tends to generate more diverse images than other models using\nreconstruction loss. Also, I figured out that the proper regularization plays a\nvital role in generative modeling.","url_abs":"http://arxiv.org/abs/1812.10650v1","url_pdf":"http://arxiv.org/pdf/1812.10650v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"sampling-using-neural-networks-for-colorizing","repo_url":"https://github.com/wayne1123/colorization","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"colorization","task_name":"Colorization"}],"methods":[{"method_slug":"colorization","method_name":"Colorization"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"cvae","method_name":"cVAE"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}