{"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/learning-to-invert-signal-recovery-via-deep","title":"Learning to Invert: Signal Recovery via Deep Convolutional Networks","arxiv_id":"1701.03891","date":"2017-01-14","proceeding":null,"authors":["Ali Mousavi","Richard G. Baraniuk"],"abstract":"The promise of compressive sensing (CS) has been offset by two significant\nchallenges. First, real-world data is not exactly sparse in a fixed basis.\nSecond, current high-performance recovery algorithms are slow to converge,\nwhich limits CS to either non-real-time applications or scenarios where massive\nback-end computing is available. In this paper, we attack both of these\nchallenges head-on by developing a new signal recovery framework we call {\\em\nDeepInverse} that learns the inverse transformation from measurement vectors to\nsignals using a {\\em deep convolutional network}. When trained on a set of\nrepresentative images, the network learns both a representation for the signals\n(addressing challenge one) and an inverse map approximating a greedy or convex\nrecovery algorithm (addressing challenge two). Our experiments indicate that\nthe DeepInverse network closely approximates the solution produced by\nstate-of-the-art CS recovery algorithms yet is hundreds of times faster in run\ntime. The tradeoff for the ultrafast run time is a computationally intensive,\noff-line training procedure typical to deep networks. However, the training\nneeds to be completed only once, which makes the approach attractive for a host\nof sparse recovery problems.","url_abs":"http://arxiv.org/abs/1701.03891v1","url_pdf":"http://arxiv.org/pdf/1701.03891v1.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":"learning-to-invert-signal-recovery-via-deep","repo_url":"https://github.com/y0umu/DeepInverse-Reimplementation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"compressive-sensing","task_name":"Compressive Sensing"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1701.03891","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}