{"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/denoising-adversarial-autoencoders","title":"Denoising Adversarial Autoencoders","arxiv_id":"1703.01220","date":"2017-03-03","proceeding":null,"authors":["Antonia Creswell","Anil Anthony Bharath"],"abstract":"Unsupervised learning is of growing interest because it unlocks the potential\nheld in vast amounts of unlabelled data to learn useful representations for\ninference. Autoencoders, a form of generative model, may be trained by learning\nto reconstruct unlabelled input data from a latent representation space. More\nrobust representations may be produced by an autoencoder if it learns to\nrecover clean input samples from corrupted ones. Representations may be further\nimproved by introducing regularisation during training to shape the\ndistribution of the encoded data in latent space. We suggest denoising\nadversarial autoencoders, which combine denoising and regularisation, shaping\nthe distribution of latent space using adversarial training. We introduce a\nnovel analysis that shows how denoising may be incorporated into the training\nand sampling of adversarial autoencoders. Experiments are performed to assess\nthe contributions that denoising makes to the learning of representations for\nclassification and sample synthesis. Our results suggest that autoencoders\ntrained using a denoising criterion achieve higher classification performance,\nand can synthesise samples that are more consistent with the input data than\nthose trained without a corruption process.","url_abs":"http://arxiv.org/abs/1703.01220v4","url_pdf":"http://arxiv.org/pdf/1703.01220v4.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":"denoising-adversarial-autoencoders","repo_url":"https://github.com/ToniCreswell/DAAE_","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"classification","task_name":"General Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1703.01220","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}