{"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/3d-shape-segmentation-with-projective","title":"3D Shape Segmentation with Projective Convolutional Networks","arxiv_id":"1612.02808","date":"2016-12-08","proceeding":"CVPR 2017 7","authors":["Evangelos Kalogerakis","Melinos Averkiou","Subhransu Maji","Siddhartha Chaudhuri"],"abstract":"This paper introduces a deep architecture for segmenting 3D objects into\ntheir labeled semantic parts. Our architecture combines image-based Fully\nConvolutional Networks (FCNs) and surface-based Conditional Random Fields\n(CRFs) to yield coherent segmentations of 3D shapes. The image-based FCNs are\nused for efficient view-based reasoning about 3D object parts. Through a\nspecial projection layer, FCN outputs are effectively aggregated across\nmultiple views and scales, then are projected onto the 3D object surfaces.\nFinally, a surface-based CRF combines the projected outputs with geometric\nconsistency cues to yield coherent segmentations. The whole architecture\n(multi-view FCNs and CRF) is trained end-to-end. Our approach significantly\noutperforms the existing state-of-the-art methods in the currently largest\nsegmentation benchmark (ShapeNet). Finally, we demonstrate promising\nsegmentation results on noisy 3D shapes acquired from consumer-grade depth\ncameras.","url_abs":"http://arxiv.org/abs/1612.02808v3","url_pdf":"http://arxiv.org/pdf/1612.02808v3.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":"3d-shape-segmentation-with-projective","repo_url":"https://github.com/kalov/ShapePFCN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"segmentation","task_name":"Segmentation"}],"methods":[{"method_slug":"crf","method_name":"CRF"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"fcn","method_name":"FCN"},{"method_slug":"max-pooling","method_name":"Max Pooling"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1612.02808","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}