{"paper":{"title":"CANSURF: An ASV-View Can Dataset and Benchmark for Detection and Tracking of Surface-Level Debris","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A dataset tailored to aluminum cans on water surfaces improves object detection accuracy twelve times over generic training sets.","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Abdullah Moosa, Mostafa Elemam, Zahra F. Rahmatullah, Zaid Aljundi","submitted_at":"2026-05-16T02:53:57Z","abstract_excerpt":"Surface-level marine debris remains a practical bottleneck for autonomous clean-up, where small, reflective targets (e.g., aluminum cans) must be detected at distance under glare, ripples, and partial submersion. This paper presents, an ASV vision system and a new surface-can dataset. The dataset comprises ~7.3k raw images extracted from videos and annotated with bounding boxes, expanded via ten augmentation types to ~57k training/validation images spanning diverse lighting and water states. A family of detector and detector-tracker pipelines tailored to surface operations were benchmarked. Tr"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Training YOLOv11 on CANSURF boosts performance 12x over generic datasets, highlighting the dataset's value.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The collected raw images and the ten augmentation types produce a training distribution that is sufficiently representative of real ASV operating conditions including glare, ripples, and partial submersion.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Presents the CANSURF dataset for surface-level aluminum can detection from ASV viewpoints and shows that training YOLOv11 on it yields a 12x performance boost over generic datasets along with stable tracking results.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A dataset tailored to aluminum cans on water surfaces improves object detection accuracy twelve times over generic training sets.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"434802bf674b2d6fd7ce949619f11fc3fc4785aee80d39b0d0dbca880c47ffec"},"source":{"id":"2605.16774","kind":"arxiv","version":1},"verdict":{"id":"54089042-ff8f-4b7f-b20a-e8f729725261","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T21:49:35.822820Z","strongest_claim":"Training YOLOv11 on CANSURF boosts performance 12x over generic datasets, highlighting the dataset's value.","one_line_summary":"Presents the CANSURF dataset for surface-level aluminum can detection from ASV viewpoints and shows that training YOLOv11 on it yields a 12x performance boost over generic datasets along with stable tracking results.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The collected raw images and the ten augmentation types produce a training distribution that is sufficiently representative of real ASV operating conditions including glare, ripples, and partial submersion.","pith_extraction_headline":"A dataset tailored to aluminum cans on water surfaces improves object detection accuracy twelve times over generic training sets."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.16774/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T22:01:19.726744Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T22:01:02.862100Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T19:01:56.307986Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.442501Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"764de35cab4da694084bc78a0922c927e5820d193423fe1eda970cb7e22f0d61"},"references":{"count":17,"sample":[{"doi":"","year":2020,"title":"Marine debris handling guide- lines,","work_id":"e3963d67-a9d1-42e1-b1bb-5799b46b718b","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"2020 international coastal cleanup: By the numbers,","work_id":"ab6728d1-223a-4b59-9986-4edb920dcf9c","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Marida: A benchmark for marine debris detection from sentinel-2 remote sensing data,","work_id":"eed024de-599f-46f3-8f00-a890957285e9","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Trash-icra19: A bounding box labeled dataset of underwater trash,","work_id":"9335cdc3-09b2-4274-b243-f4652e1ad146","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Trashcan 1.0: An instance- segmentation labeled dataset of trash observations,","work_id":"744fe0a3-14eb-4dc9-9902-3d2858334c4b","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":17,"snapshot_sha256":"78dfb717d6b0f600c0feaa8f76ef063101986a870ed5e4729f336d723f080bbe","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"}