{"id":"3fb71384-65aa-44f5-ae29-d406a68ffedf","slug":"onecanvas","name":"OneCanvas","description":"OneCanvas (3D Scene Understanding via Panoramic Reprojection) is a 2026 research method from Technical University of Munich and Huawei (Baranowski, Chen, Nießner; NeurIPS 2026 Spotlight) that lifts multi-view RGB-D patch features onto one equirectangular panoramic canvas so a frozen 2D VLM can do 3D scene understanding and situated reasoning with far less training compute than geometry-encoder or large-data alternatives.","brainType":"research-model","isOpen":false,"maturityStage":"research","architecture":"Frozen VLM vision encoder (Qwen3-VL-8B class) extracts per-view patches; depth + pose lift to 3D; continuous equirectangular canvas tokens with native MRoPE (longitude/latitude/frame) plus additive metric 3D position embedding; LoRA adapters with two-stage spatial curriculum then target QA adaptation.","reviewStatus":"reviewed","sources":[{"url":"https://baranowskibrt.github.io/onecanvas/","title":"OneCanvas: 3D Scene Understanding via Panoramic Reprojection","sourceName":"OneCanvas project page","publishedAt":"2026"},{"url":"https://arxiv.org/abs/2606.19253","title":"OneCanvas: 3D Scene Understanding via Panoramic Reprojection (arXiv:2606.19253)","sourceName":"arXiv","publishedAt":"2026"},{"url":"https://x.com/robotsdigest/status/2106510045513154969","title":"Robots Digest: OneCanvas summary","sourceName":"X (@robotsdigest)","publishedAt":"2026-10-03"}],"keyFacts":[{"label":"SQA3D 65.3 EM@1 / VSI-Bench 71.3 / SPBench 72.1 zero-shot (research-stated)","value":"Research-stated (research-stated; also arXiv; Robots Digest X): OneCanvas (Baranowski / Chen / Nießner; TUM + Huawei; NeurIPS 2026 Spotlight) reports SOTA spatial reasoning on SQA3D (65.3 EM@1, +2.3 over prior best), VSI-Bench (71.3 average), and SPBench (72.1 zero-shot overall, +4.8), claiming an order of magnitude less training compute than closest competitors. Cap: author project-page / paper tables; not an independent replication."},{"label":"Panoramic patch-feature canvas for frozen 2D VLMs (research-stated)","value":"Research-stated (research-stated; also arXiv): method lifts multi-view patch features into 3D via depth and camera pose, places each patch as a continuous token on one equirectangular panorama (longitude/latitude from a chosen origin), adds a metric 3D position embedding, and feeds a pretrained VLM (Qwen3-VL-8B class) with LoRA adapters only. Cap: author method description."}],"aliases":["OneCanvas","One Canvas","OneCanvas panoramic reprojection"],"collisionRisk":"low","reviewNote":null,"builtOnBrainId":null,"createdAt":"2026-10-04T01:50:04.501Z","updatedAt":"2026-10-10T21:05:52.525Z","jsonLd":{"@context":"https://schema.org","@type":"SoftwareApplication","@id":"https://registry.deploy.report/brains/onecanvas","url":"https://registry.deploy.report/brains/onecanvas","name":"OneCanvas","alternateName":["OneCanvas","One Canvas","OneCanvas panoramic reprojection"],"description":"OneCanvas (3D Scene Understanding via Panoramic Reprojection) is a 2026 research method from Technical University of Munich and Huawei (Baranowski, Chen, Nießner; NeurIPS 2026 Spotlight) that lifts multi-view RGB-D patch features onto one equirectangular panoramic canvas so a frozen 2D VLM can do 3D scene understanding and situated reasoning with far less training compute than geometry-encoder or large-data alternatives.","identifier":"3fb71384-65aa-44f5-ae29-d406a68ffedf","applicationCategory":"research-model","publisher":{"@id":"https://deploy.report/#organization"}},"framework_metadata":{"framework_schema_version":"0.1.0","verification_status":"verified","maturity_stage":"research","lifecycle_state":null,"architectural_position":{"cohort":null,"sub_cohorts":[]},"within_cohort_verified_vs_claimed_pair":null,"cap_flags":[],"verification_depth":{"sources_count":3,"primary_source_types":["preprint"]}}}