{"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/an-efficient-deep-convolutional-laplacian","title":"An efficient deep convolutional laplacian pyramid architecture for CS reconstruction at low sampling ratios","arxiv_id":"1804.04970","date":"2018-04-13","proceeding":null,"authors":["Wenxue Cui","Heyao Xu","Xinwei Gao","Shengping Zhang","Feng Jiang","Debin Zhao"],"abstract":"The compressed sensing (CS) has been successfully applied to image\ncompression in the past few years as most image signals are sparse in a certain\ndomain. Several CS reconstruction models have been proposed and obtained\nsuperior performance. However, these methods suffer from blocking artifacts or\nringing effects at low sampling ratios in most cases. To address this problem,\nwe propose a deep convolutional Laplacian Pyramid Compressed Sensing Network\n(LapCSNet) for CS, which consists of a sampling sub-network and a\nreconstruction sub-network. In the sampling sub-network, we utilize a\nconvolutional layer to mimic the sampling operator. In contrast to the fixed\nsampling matrices used in traditional CS methods, the filters used in our\nconvolutional layer are jointly optimized with the reconstruction sub-network.\nIn the reconstruction sub-network, two branches are designed to reconstruct\nmulti-scale residual images and muti-scale target images progressively using a\nLaplacian pyramid architecture. The proposed LapCSNet not only integrates\nmulti-scale information to achieve better performance but also reduces\ncomputational cost dramatically. Experimental results on benchmark datasets\ndemonstrate that the proposed method is capable of reconstructing more details\nand sharper edges against the state-of-the-arts methods.","url_abs":"http://arxiv.org/abs/1804.04970v1","url_pdf":"http://arxiv.org/pdf/1804.04970v1.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":"an-efficient-deep-convolutional-laplacian","repo_url":"https://github.com/WenxueCui/LapCSNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"blocking","task_name":"Blocking"},{"task_slug":"compressive-sensing","task_name":"Compressive Sensing"},{"task_slug":"image-compression","task_name":"Image Compression"},{"task_slug":"compressed-sensing","task_name":"compressed sensing"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/compressive-sensing-on-set5","task":"Compressive Sensing","dataset":"Set5","model":"LapCSNet","rank_in_archive_order":1,"of":1,"metrics":{"Average PSNR":"improving about 0.2-0.3dB"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}