{"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/deeplidar-deep-surface-normal-guided-depth","title":"DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene from Sparse LiDAR Data and Single Color Image","arxiv_id":"1812.00488","date":"2018-12-02","proceeding":"CVPR 2019 6","authors":["Jiaxiong Qiu","Zhaopeng Cui","yinda zhang","Xingdi Zhang","Shuaicheng Liu","Bing Zeng","Marc Pollefeys"],"abstract":"In this paper, we propose a deep learning architecture that produces accurate\ndense depth for the outdoor scene from a single color image and a sparse depth.\nInspired by the indoor depth completion, our network estimates surface normals\nas the intermediate representation to produce dense depth, and can be trained\nend-to-end. With a modified encoder-decoder structure, our network effectively\nfuses the dense color image and the sparse LiDAR depth. To address outdoor\nspecific challenges, our network predicts a confidence mask to handle mixed\nLiDAR signals near foreground boundaries due to occlusion, and combines\nestimates from the color image and surface normals with learned attention maps\nto improve the depth accuracy especially for distant areas. Extensive\nexperiments demonstrate that our model improves upon the state-of-the-art\nperformance on KITTI depth completion benchmark. Ablation study shows the\npositive impact of each model components to the final performance, and\ncomprehensive analysis shows that our model generalizes well to the input with\nhigher sparsity or from indoor scenes.","url_abs":"http://arxiv.org/abs/1812.00488v2","url_pdf":"http://arxiv.org/pdf/1812.00488v2.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":"deeplidar-deep-surface-normal-guided-depth","repo_url":"https://github.com/JiaxiongQ/DeepLiDAR","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"depth-completion","task_name":"Depth Completion"},{"task_slug":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"depth-prediction","task_name":"Depth Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1812.00488","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}