{"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/laplacian-pyramid-reconstruction-and","title":"Laplacian Pyramid Reconstruction and Refinement for Semantic Segmentation","arxiv_id":"1605.02264","date":"2016-05-08","proceeding":null,"authors":["Golnaz Ghiasi","Charless C. Fowlkes"],"abstract":"CNN architectures have terrific recognition performance but rely on spatial\npooling which makes it difficult to adapt them to tasks that require dense,\npixel-accurate labeling. This paper makes two contributions: (1) We demonstrate\nthat while the apparent spatial resolution of convolutional feature maps is\nlow, the high-dimensional feature representation contains significant sub-pixel\nlocalization information. (2) We describe a multi-resolution reconstruction\narchitecture based on a Laplacian pyramid that uses skip connections from\nhigher resolution feature maps and multiplicative gating to successively refine\nsegment boundaries reconstructed from lower-resolution maps. This approach\nyields state-of-the-art semantic segmentation results on the PASCAL VOC and\nCityscapes segmentation benchmarks without resorting to more complex\nrandom-field inference or instance detection driven architectures.","url_abs":"http://arxiv.org/abs/1605.02264v2","url_pdf":"http://arxiv.org/pdf/1605.02264v2.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":"laplacian-pyramid-reconstruction-and","repo_url":"https://github.com/golnazghiasi/LRR","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/semantic-segmentation-on-cityscapes","task":"Semantic Segmentation","dataset":"Cityscapes test","model":"LRR-4x","rank_in_archive_order":72,"of":105,"metrics":{"Mean IoU (class)":"71.8%"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1605.02264","atlas_url":"https://app.syntology.ai/?focus=1605.02264","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}