{"paper":{"title":"OptMap: Geometric Map Distillation via Submodular Maximization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"OptMap distills large LiDAR streams into compact application-specific maps by maximizing a submodular reward function with polynomial-time near-optimal algorithms.","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Brett T. Lopez, Christa S. Robison, David Thorne, Nathan Chan, Philip R. Osteen","submitted_at":"2025-12-08T17:56:57Z","abstract_excerpt":"Autonomous robots rely on geometric maps to inform a diverse set of perception and decision-making algorithms. As autonomy requires reasoning and planning on multiple scales, each algorithm may require a different map for optimal performance. LiDAR sensors generate an abundance of geometric data (up to 50 MB per second) to satisfy these diverse requirements. However, the point-based operations required to process perception data are both memory and computationally expensive. Such operations can be bypassed via learned representations that encode similarity, but selecting informative, size-cons"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We present OptMap: a geometric map distillation algorithm which achieves online, application-specific map generation via multiple theoretical and algorithmic innovations. A central feature is the maximization of set functions that exhibit diminishing returns, i.e., submodularity, using polynomial-time algorithms with provably near-optimal solutions.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the proposed reward function is submodular (or sufficiently close) so that the polynomial-time greedy-style algorithms retain their near-optimality guarantees when applied to real LiDAR streams.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"OptMap generates compact, application-specific geometric maps from streaming LiDAR data using a novel submodular reward function and a dynamically reordered streaming maximization algorithm.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"OptMap distills large LiDAR streams into compact application-specific maps by maximizing a submodular reward function with polynomial-time near-optimal algorithms.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"48df3fec5ce2a5c4f418e8ef381fcf213bfbc6239b05faa38769107811ae5259"},"source":{"id":"2512.07775","kind":"arxiv","version":2},"verdict":{"id":"404297c2-02b8-4b5e-a72b-6da31eedf6be","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T00:29:11.063863Z","strongest_claim":"We present OptMap: a geometric map distillation algorithm which achieves online, application-specific map generation via multiple theoretical and algorithmic innovations. A central feature is the maximization of set functions that exhibit diminishing returns, i.e., submodularity, using polynomial-time algorithms with provably near-optimal solutions.","one_line_summary":"OptMap generates compact, application-specific geometric maps from streaming LiDAR data using a novel submodular reward function and a dynamically reordered streaming maximization algorithm.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the proposed reward function is submodular (or sufficiently close) so that the polynomial-time greedy-style algorithms retain their near-optimality guarantees when applied to real LiDAR streams.","pith_extraction_headline":"OptMap distills large LiDAR streams into compact application-specific maps by maximizing a submodular reward function with polynomial-time near-optimal algorithms."},"references":{"count":61,"sample":[{"doi":"","year":2022,"title":"Overlaptransformer: An efficient and yaw-angle-invariant transformer network for lidar-based place recognition,","work_id":"7e4603dc-620b-403c-850e-5bdbcbc6a155","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Overlapnet: A siamese network for computing lidar scan similarity with applications to loop closing and localization,","work_id":"48c69809-24cc-4589-a88e-05c9e3e82cc5","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Padloc: Lidar-based deep loop closure detection and registration using panoptic attention,","work_id":"3cb732ac-ac60-47a8-b63e-aa4e9ca3bbc8","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"Minkloc3d: Point cloud based large-scale place recog- nition,","work_id":"523ff79a-c759-4954-8b49-d6cd0be4efc0","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Submodular optimization for keyframe selection & usage in slam,","work_id":"1f9dd0ce-7069-4671-9e2c-1879d9bfcac5","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":61,"snapshot_sha256":"fd2bac47905bb00aea4675fe0023d6996b0c24ba7586a4b68d09c3ff668e1cac","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"be25867fc8649f07a2a4d82ec2f714d297a271732d49b728c96c4c1258c1b237"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}