{"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/multi-view-silhouette-and-depth-decomposition","title":"Multi-View Silhouette and Depth Decomposition for High Resolution 3D Object Representation","arxiv_id":"1802.09987","date":"2018-02-27","proceeding":"NeurIPS 2018 12","authors":["Edward Smith","Scott Fujimoto","David Meger"],"abstract":"We consider the problem of scaling deep generative shape models to\nhigh-resolution. Drawing motivation from the canonical view representation of\nobjects, we introduce a novel method for the fast up-sampling of 3D objects in\nvoxel space through networks that perform super-resolution on the six\northographic depth projections. This allows us to generate high-resolution\nobjects with more efficient scaling than methods which work directly in 3D. We\ndecompose the problem of 2D depth super-resolution into silhouette and depth\nprediction to capture both structure and fine detail. This allows our method to\ngenerate sharp edges more easily than an individual network. We evaluate our\nwork on multiple experiments concerning high-resolution 3D objects, and show\nour system is capable of accurately predicting novel objects at resolutions as\nlarge as 512$\\mathbf{\\times}$512$\\mathbf{\\times}$512 -- the highest resolution\nreported for this task. We achieve state-of-the-art performance on 3D object\nreconstruction from RGB images on the ShapeNet dataset, and further demonstrate\nthe first effective 3D super-resolution method.","url_abs":"http://arxiv.org/abs/1802.09987v3","url_pdf":"http://arxiv.org/pdf/1802.09987v3.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":"multi-view-silhouette-and-depth-decomposition","repo_url":"https://github.com/EdwardSmith1884/3D-Object-Super-Resolution","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multi-view-silhouette-and-depth-decomposition","repo_url":"https://github.com/EdwardSmith1884/Multi-View-Silhouette-and-Depth-Decomposition-for-High-Resolution-3D-Object-Representation","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multi-view-silhouette-and-depth-decomposition","repo_url":"https://github.com/kingcheng2000/Multi-View-Silhouette-and-Depth-Decomposition-for-High-Resolution-3D-Object-Representation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"3d-object-reconstruction","task_name":"3D Object Reconstruction"},{"task_slug":"3d-object-super-resolution","task_name":"3D Object Super-Resolution"},{"task_slug":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"depth-prediction","task_name":"Depth Prediction"},{"task_slug":"object-reconstruction","task_name":"Object Reconstruction"},{"task_slug":"super-resolution","task_name":"Super-Resolution"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/3d-object-reconstruction-on-data3dr2n2","task":"3D Object Reconstruction","dataset":"Data3D−R2N2","model":"MVD","rank_in_archive_order":11,"of":15,"metrics":{"Avg F1":"66.39"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1802.09987","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}