{"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/ista-net-interpretable-optimization-inspired","title":"ISTA-Net: Interpretable Optimization-Inspired Deep Network for Image Compressive Sensing","arxiv_id":"1706.07929","date":"2017-06-24","proceeding":"CVPR 2018 6","authors":["Jian Zhang","Bernard Ghanem"],"abstract":"With the aim of developing a fast yet accurate algorithm for compressive\nsensing (CS) reconstruction of natural images, we combine in this paper the\nmerits of two existing categories of CS methods: the structure insights of\ntraditional optimization-based methods and the speed of recent network-based\nones. Specifically, we propose a novel structured deep network, dubbed\nISTA-Net, which is inspired by the Iterative Shrinkage-Thresholding Algorithm\n(ISTA) for optimizing a general $\\ell_1$ norm CS reconstruction model. To cast\nISTA into deep network form, we develop an effective strategy to solve the\nproximal mapping associated with the sparsity-inducing regularizer using\nnonlinear transforms. All the parameters in ISTA-Net (\\eg nonlinear transforms,\nshrinkage thresholds, step sizes, etc.) are learned end-to-end, rather than\nbeing hand-crafted. Moreover, considering that the residuals of natural images\nare more compressible, an enhanced version of ISTA-Net in the residual domain,\ndubbed {ISTA-Net}$^+$, is derived to further improve CS reconstruction.\nExtensive CS experiments demonstrate that the proposed ISTA-Nets outperform\nexisting state-of-the-art optimization-based and network-based CS methods by\nlarge margins, while maintaining fast computational speed. Our source codes are\navailable: \\textsl{http://jianzhang.tech/projects/ISTA-Net}.","url_abs":"http://arxiv.org/abs/1706.07929v2","url_pdf":"http://arxiv.org/pdf/1706.07929v2.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":"ista-net-interpretable-optimization-inspired","repo_url":"https://github.com/jianzhangcs/ISTA-Net-PyTorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"compressive-sensing","task_name":"Compressive Sensing"}],"methods":[{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}