{"ID":5675162,"CreatedAt":"2026-07-03T01:40:09.565152011Z","UpdatedAt":"2026-07-05T06:17:44.354484894Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.01757","arxiv_id":"2607.01757","title":"DL-VINS-Factory: A Modular Framework for Learned Visual Front-Ends in Visual-Inertial SLAM","abstract":"Deep-learning features excel in visual matching, yet their practical value in tightly coupled visual-inertial SLAM (VI-SLAM) remains insufficiently characterized. We present DL-VINS-Factory, a unified framework that integrates learned feature extractors (ALIKED, RaCo, SuperPoint, XFeat) with either Lucas--Kanade (LK) optical-flow tracking or LightGlue (LG) descriptor matching. All front-ends share a sliding-window Ceres back-end, with optional AnyLoc DINOv2-VLAD loop closure, and 4-DoF pose-graph optimization. We benchmark the system across the four datasets covering indoor, unstructured outdoor, aggressive-motion, and visually degraded conditions. Results show that learned front-ends are viable for real-time embedded VI-SLAM, but are not universally superior to classical tracking. Relative to the corresponding GFTT+LK baseline, ALIKED+LG reduces EuRoC ATE by $5\\%$ in monocular odometry and by $7\\%$ in stereo with loop-closure. On NTU-VIRAL, where aggressive aerial motion increases inter-frame viewpoint change, ALIKED+LG stereo reduces loop-closed ATE by $12\\%$. In Botanic Garden dataset, optical-flow tracking remains preferable, but learned keypoints still improve over the baseline GFTT, in which SuperPoint+LK reduces grayscale camera ATE by $29\\%$, while RaCo+LK reduces RGB camera ATE by $38\\%$. On SubT-MRS, learned front-ends display varying degree of improvement based on individual cases. With TensorRT acceleration on a Jetson AGX Orin, all valid configurations run in real time between $29$--$47$ FPS in monocular mode and $18$--$33$ FPS in stereo mode for the EuRoC and NTU-VIRAL datasets. AnyLoc further confirms roughly $2$--$7\\times$ more valid loops than BRIEF+DBoW2. The implementation is open-sourced at https://github.com/limshoonkit/DL-VINS-Factory-ROS2/.","short_abstract":"Deep-learning features excel in visual matching, yet their practical value in tightly coupled visual-inertial SLAM (VI-SLAM) remains insufficiently characterized. We present DL-VINS-Factory, a unified framework that integrates learned feature extractors (ALIKED, RaCo, SuperPoint, XFeat) with either Lucas--Kanade (LK) o...","url_abs":"https://arxiv.org/abs/2607.01757","url_pdf":"https://arxiv.org/pdf/2607.01757v1","authors":"[\"Shoon Kit Lim\",\"Melissa Jia Ying Chong\",\"Ting Yang Ling\"]","published":"2026-07-02T06:17:33Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.RO\"]","methods":"[]","has_code":false,"code_links":[{"ID":613880,"CreatedAt":"2026-07-03T01:40:09.565152011Z","UpdatedAt":"2026-07-03T01:40:09.565152011Z","DeletedAt":null,"paper_id":5675162,"paper_url":"https://arxiv.org/abs/2607.01757","paper_title":"DL-VINS-Factory: A Modular Framework for Learned Visual Front-Ends in Visual-Inertial SLAM","repo_url":"https://github.com/limshoonkit/DL-VINS-Factory-ROS2","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}