{"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/cnn-svo-improving-the-mapping-in-semi-direct","title":"CNN-SVO: Improving the Mapping in Semi-Direct Visual Odometry Using Single-Image Depth Prediction","arxiv_id":"1810.01011","date":"2018-10-01","proceeding":null,"authors":["Shing Yan Loo","Ali Jahani Amiri","Syamsiah Mashohor","Sai Hong Tang","Hong Zhang"],"abstract":"Reliable feature correspondence between frames is a critical step in visual\nodometry (VO) and visual simultaneous localization and mapping (V-SLAM)\nalgorithms. In comparison with existing VO and V-SLAM algorithms, semi-direct\nvisual odometry (SVO) has two main advantages that lead to state-of-the-art\nframe rate camera motion estimation: direct pixel correspondence and efficient\nimplementation of probabilistic mapping method. This paper improves the SVO\nmapping by initializing the mean and the variance of the depth at a feature\nlocation according to the depth prediction from a single-image depth prediction\nnetwork. By significantly reducing the depth uncertainty of the initialized map\npoint (i.e., small variance centred about the depth prediction), the benefits\nare twofold: reliable feature correspondence between views and fast convergence\nto the true depth in order to create new map points. We evaluate our method\nwith two outdoor datasets: KITTI dataset and Oxford Robotcar dataset. The\nexperimental results indicate that the improved SVO mapping results in\nincreased robustness and camera tracking accuracy.","url_abs":"http://arxiv.org/abs/1810.01011v1","url_pdf":"http://arxiv.org/pdf/1810.01011v1.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":"cnn-svo-improving-the-mapping-in-semi-direct","repo_url":"https://github.com/yan99033/CNN-SVO","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"cnn-svo-improving-the-mapping-in-semi-direct","repo_url":"https://github.com/muskie82/CNN-DSO","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"depth-prediction","task_name":"Depth Prediction"},{"task_slug":"motion-estimation","task_name":"Motion Estimation"},{"task_slug":"prediction","task_name":"Prediction"},{"task_slug":"simultaneous-localization-and-mapping","task_name":"Simultaneous Localization and Mapping"},{"task_slug":"visual-odometry","task_name":"Visual Odometry"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1810.01011","atlas_url":"https://app.syntology.ai/?focus=1810.01011","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}