{"paper":{"title":"Argoverse 2: Next Generation Datasets for Self-Driving Perception and Forecasting","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Argoverse 2 releases three large datasets to support new research in self-driving perception and forecasting.","cross_cats":["cs.AI","cs.LG","cs.RO"],"primary_cat":"cs.CV","authors_text":"Andrew Hartnett, Benjamin Wilson, Bowen Pan, Deva Ramanan, Jagjeet Singh, James Hays, Jhony Kaesemodel Pontes, John Lambert, Peter Carr, Ratnesh Kumar, Siddhesh Khandelwal, Tanmay Agarwal, William Qi","submitted_at":"2023-01-02T00:36:22Z","abstract_excerpt":"We introduce Argoverse 2 (AV2) - a collection of three datasets for perception and forecasting research in the self-driving domain. The annotated Sensor Dataset contains 1,000 sequences of multimodal data, encompassing high-resolution imagery from seven ring cameras, and two stereo cameras in addition to lidar point clouds, and 6-DOF map-aligned pose. Sequences contain 3D cuboid annotations for 26 object categories, all of which are sufficiently-sampled to support training and evaluation of 3D perception models. The Lidar Dataset contains 20,000 sequences of unlabeled lidar point clouds and ma"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We believe these datasets will support new and existing machine learning research problems in ways that existing datasets do not.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the provided annotations are accurate enough and the selected scenarios sufficiently representative to drive meaningful improvements in deployed self-driving systems.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Argoverse 2 introduces three new datasets with annotated sensor data, massive lidar collections, and challenging motion forecasting scenarios for autonomous driving research.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Argoverse 2 releases three large datasets to support new research in self-driving perception and forecasting.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"c1aebfb621b3bb1bbe9c690e6b34ff203451091304b61e6561bfb938ea30506c"},"source":{"id":"2301.00493","kind":"arxiv","version":1},"verdict":{"id":"077689b3-084e-429b-bf06-0e62bc5ebfd6","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-12T20:10:13.850741Z","strongest_claim":"We believe these datasets will support new and existing machine learning research problems in ways that existing datasets do not.","one_line_summary":"Argoverse 2 introduces three new datasets with annotated sensor data, massive lidar collections, and challenging motion forecasting scenarios for autonomous driving research.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the provided annotations are accurate enough and the selected scenarios sufficiently representative to drive meaningful improvements in deployed self-driving systems.","pith_extraction_headline":"Argoverse 2 releases three large datasets to support new research in self-driving perception and forecasting."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2301.00493/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":54,"sample":[{"doi":"","year":2019,"title":"SemanticKITTI: A dataset for semantic scene understanding of lidar sequences","work_id":"27a8f474-426a-44fe-84d3-baa5116bce2d","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Range conditioned dilated convolutions for scale invariant 3d object detection","work_id":"99a248a9-df28-4e84-9041-34ce5b65aced","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2005,"title":"Language Models are Few-Shot Learners","work_id":"214732c0-2edd-44a0-af9e-28184a2b8279","ref_index":3,"cited_arxiv_id":"2005.14165","is_internal_anchor":true},{"doi":"","year":2020,"title":"Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Giancarlo Baldan, and Oscar Beijbom","work_id":"09126d7a-6b02-4972-b45c-7ff87220ed10","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"To the point: Efficient 3d object detection in the range image with graph convolution kernels","work_id":"a733b164-6e16-483c-ac28-24feabff3117","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":54,"snapshot_sha256":"8b49227c34dae52f88f7ca1bb866e3007c8c1293c2ac943d038693752d37418f","internal_anchors":2},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}