{"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/variational-approaches-for-auto-encoding","title":"Variational Approaches for Auto-Encoding Generative Adversarial Networks","arxiv_id":"1706.04987","date":"2017-06-15","proceeding":null,"authors":["Mihaela Rosca","Balaji Lakshminarayanan","David Warde-Farley","Shakir Mohamed"],"abstract":"Auto-encoding generative adversarial networks (GANs) combine the standard GAN\nalgorithm, which discriminates between real and model-generated data, with a\nreconstruction loss given by an auto-encoder. Such models aim to prevent mode\ncollapse in the learned generative model by ensuring that it is grounded in all\nthe available training data. In this paper, we develop a principle upon which\nauto-encoders can be combined with generative adversarial networks by\nexploiting the hierarchical structure of the generative model. The underlying\nprinciple shows that variational inference can be used a basic tool for\nlearning, but with the in- tractable likelihood replaced by a synthetic\nlikelihood, and the unknown posterior distribution replaced by an implicit\ndistribution; both synthetic likelihoods and implicit posterior distributions\ncan be learned using discriminators. This allows us to develop a natural fusion\nof variational auto-encoders and generative adversarial networks, combining the\nbest of both these methods. We describe a unified objective for optimization,\ndiscuss the constraints needed to guide learning, connect to the wide range of\nexisting work, and use a battery of tests to systematically and quantitatively\nassess the performance of our method.","url_abs":"http://arxiv.org/abs/1706.04987v2","url_pdf":"http://arxiv.org/pdf/1706.04987v2.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":"variational-approaches-for-auto-encoding","repo_url":"https://github.com/kryvosheyev/xray-anomaly-detection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"variational-approaches-for-auto-encoding","repo_url":"https://github.com/lkhphuc/Anomaly-BiGAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"variational-approaches-for-auto-encoding","repo_url":"https://github.com/lkhphuc/Anomaly-XRay-GANs","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"variational-approaches-for-auto-encoding","repo_url":"https://github.com/pavasgdb/Anomaly-detector-using-GAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"variational-approaches-for-auto-encoding","repo_url":"https://github.com/zzmtsvv/adversarial","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"variational-approaches-for-auto-encoding","repo_url":"https://github.com/PrateekMunjal/Alpha_GAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"variational-inference","task_name":"Variational Inference"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1706.04987","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}