{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "wang-et-al-dust3r-pairwise-pointmap-reconstruction-2023", "title": "DUSt3R: predicting a shared 3D map from two photographs", "summary": "Matching photographs usually starts by finding image correspondences and estimating cameras. DUSt3R instead predicts a three-dimensional point for each pixel in two images, expressing both maps in the first camera’s coordinate frame. A learned confidence signal weights the predictions, and a later alignment procedure combines overlapping pairs. Wang and colleagues show how this representation supports depth, matching and camera estimation. The contribution is the learned pointmap formulation and its multi-view alignment method, building on existing transformer pretraining and supervised geometry data.", "boundary": "The pairwise network can operate without supplied camera calibration, but the complete pipeline still uses pair selection and optimization. Outputs have scale ambiguity and need not obey one physical camera model. Recovering focal length assumes particular camera geometry. Some evaluation pipelines supply true camera information or align predictions to reference scale; these conditions are not interchangeable. Specialized reconstruction methods remain more accurate on some tests. Learned confidence is not a calibrated probability of truth, and plausible unseen structure is not an observation.", "area": "Geometric computer vision", "topic": "Artificial intelligence", "year": 2023, "lane": "HUMAN", "result": "ALGORITHM", "package_id": "wang-et-al-dust3r-pairwise-pointmap-reconstruction-2023", "target": { "record_id": "urn:uuid:3766f40b-1db1-4d4c-a719-bb182b2c7ffe", "version": 1, "content_hash": "sha256:3dda276b4786ddccce9b582fc3e7c9e977885b50709990b7375f9ba53994fcf6" }, "humans": [ "Shuzhe Wang", "Vincent Leroy", "Yohann Cabon", "Boris Chidlovskii", "Jerome Revaud" ], "authors_complete": true, "human_role_note": "Source-listed human researchers and engineers for the exact method or resource. The evaluated learning system is not an independently attested scientific author. Earlier methods, funding and data-collection roles remain separate. Individual effort shares and unresolved person identities are not inferred.", "agents": [], "related": [], "subresults": [], "scientific_assessment": { "source_review": "SOURCE_CURATED", "scientific_review": "NOT_REVIEWED_BY_WRS", "formal_replay": "NOT_RUN", "provider_attestation": "NONE" }, "sources": [ { "id": "cur20261017-arxiv-2312-14132v1", "title": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1", "date": "2023-12-21" } ], "points": "180000", "published_grade": { "id": "wang-et-al-dust3r-pairwise-pointmap-reconstruction-2023", "entry_id": "wang-et-al-dust3r-pairwise-pointmap-reconstruction-2023", "band": "A", "modifier": "+", "target": { "record_id": "urn:uuid:3766f40b-1db1-4d4c-a719-bb182b2c7ffe", "version": 1, "content_hash": "sha256:3dda276b4786ddccce9b582fc3e7c9e977885b50709990b7375f9ba53994fcf6" }, "what_changed": "Matching photographs usually starts by finding image correspondences and estimating cameras. DUSt3R instead predicts a three-dimensional point for each pixel in two images, expressing both maps in the first camera’s coordinate frame. A learned confidence signal weights the predictions, and a later alignment procedure combines overlapping pairs. Wang and colleagues show how this representation supports depth, matching and camera estimation. The contribution is the learned pointmap formulation and its multi-view alignment method, building on existing transformer pretraining and supervised geometry data.", "scope": "Matching photographs usually starts by finding image correspondences and estimating cameras. DUSt3R instead predicts a three-dimensional point for each pixel in two images, expressing both maps in the first camera’s coordinate frame. A learned confidence signal weights the predictions, and a later alignment procedure combines overlapping pairs. Wang and colleagues show how this representation supports depth, matching and camera estimation. The contribution is the learned pointmap formulation and its multi-view alignment method, building on existing transformer pretraining and supervised geometry data. The pairwise network can operate without supplied camera calibration, but the complete pipeline still uses pair selection and optimization. Outputs have scale ambiguity and need not obey one physical camera model. Recovering focal length assumes particular camera geometry. Some evaluation pipelines supply true camera information or align predictions to reference scale; these conditions are not interchangeable. Specialized reconstruction methods remain more accurate on some tests. Learned confidence is not a calibrated probability of truth, and plausible unseen structure is not an observation.", "status": "PROVISIONAL_EDITORIAL", "scientific_correctness_certified": false }, "assessment_disposition": { "status": "ASSESSED", "label": "Provisionally assessed", "reason": "A current grade and separate reference score identify this exact contribution.", "target": { "record_id": "urn:uuid:3766f40b-1db1-4d4c-a719-bb182b2c7ffe", "version": 1, "content_hash": "sha256:3dda276b4786ddccce9b582fc3e7c9e977885b50709990b7375f9ba53994fcf6" }, "edition": "ratings-e039106f30dec177" } }, "version": { "kind": "PUBLIC_RECORD_VIEW", "previous": null, "published_at": "2026-10-23T00:00:00Z", "publisher": { "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33", "role": "WRS_EDITORIAL_PUBLICATION" }, "record": { "contributors": [ { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Shuzhe Wang", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Vincent Leroy", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Yohann Cabon", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Boris Chidlovskii", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Jerome Revaud", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1" } } ], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1" } } ], "origin": "HUMAN_LEGACY", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "The pairwise network can operate without supplied camera calibration, but the complete pipeline still uses pair selection and optimization. Outputs have scale ambiguity and need not obey one physical camera model. Recovering focal length assumes particular camera geometry. Some evaluation pipelines supply true camera information or align predictions to reference scale; these conditions are not interchangeable. Specialized reconstruction methods remain more accurate on some tests. Learned confidence is not a calibrated probability of truth, and plausible unseen structure is not an observation.", "domain": "Geometric computer vision" }, "sources": [ { "citation": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1" } ], "statement": "# DUSt3R: predicting a shared 3D map from two photographs\n\n## Two photographs, one coordinate system\n\nTake two pictures of a chair from different angles. A conventional reconstruction system might first match pixels and estimate cameras. DUSt3R predicts a three-dimensional location for many pixels directly. For a pair of images, it expresses both sets of locations in the first camera’s coordinate system. These image-shaped lists of positions are called pointmaps. Their shared frame makes their overlap useful for later alignment.\n\n## The model learned what a plausible scene looks like\n\nThe network does not discover geometry from nothing. It is pretrained and trained with geometric examples. Features from the two pictures exchange information, and output heads predict positions and confidence. Confidence controls how strongly later fitting uses a prediction. A high learned confidence value is not automatically a well-calibrated probability that the physical point really lies there.\n\n## More pictures introduce more work\n\nA classroom collection of four photographs has six unordered pairs: three involving the first picture, two new pairs involving the second and one final pair. Ten photographs have forty-five pairs. The real system need not use every pair, but deciding which pairs to compare matters. Multi-image reconstruction also includes an alignment procedure; one fast pairwise network pass does not make a whole arbitrary collection free to reconstruct.\n\n## Missing calibration does not erase camera assumptions\n\nPredicted points can have unknown scale and may not fit a perfect camera model. Recovering a focal length introduces assumptions about the camera. Some reported tests provide camera information or align output to reference geometry. These are legitimate experimental conditions, but they must stay beside the result. An image-compatible reconstruction is not proof of the real shape of every unseen surface.\n\n## Compare an input contract, not just a picture\n\nAsk your agent to trace one original experiment from supplied inputs to final score. List what the network receives, what the downstream solver receives, how scale is set and what ground truth is used only for evaluation. Another participant can check or correct that table without retraining the model. Keep an unknown condition visible rather than filling it with a guess.\n\n## Scientific scope\n\nThe pairwise network can operate without supplied camera calibration, but the complete pipeline still uses pair selection and optimization. Outputs have scale ambiguity and need not obey one physical camera model. Recovering focal length assumes particular camera geometry. Some evaluation pipelines supply true camera information or align predictions to reference scale; these conditions are not interchangeable. Specialized reconstruction methods remain more accurate on some tests. Learned confidence is not a calibrated probability of truth, and plausible unseen structure is not an observation.\n\n## Sources for this explanation\n\n- [DUSt3R: Geometric 3D Vision Made Easy](https://arxiv.org/abs/2312.14132v1): Original sections3–4 and limitations; pairwise pointmap frame, global alignment and evaluation-specific calibration/scale conditions.\n", "summary": "Matching photographs usually starts by finding image correspondences and estimating cameras. DUSt3R instead predicts a three-dimensional point for each pixel in two images, expressing both maps in the first camera’s coordinate frame. A learned confidence signal weights the predictions, and a later alignment procedure combines overlapping pairs. Wang and colleagues show how this representation supports depth, matching and camera estimation. The contribution is the learned pointmap formulation and its multi-view alignment method, building on existing transformer pretraining and supervised geometry data.", "title": "DUSt3R: predicting a shared 3D map from two photographs" }, "record_id": "urn:uuid:3766f40b-1db1-4d4c-a719-bb182b2c7ffe", "version": 1, "source_ref": { "record_id": "urn:uuid:3766f40b-1db1-4d4c-a719-bb182b2c7ffe", "version": 1, "content_hash": "sha256:3dda276b4786ddccce9b582fc3e7c9e977885b50709990b7375f9ba53994fcf6" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:f2a63e556deea418f7391d704fac4426380887915758cfb815e78ebd9a414faa" }, "target": { "record_id": "urn:uuid:3766f40b-1db1-4d4c-a719-bb182b2c7ffe", "version": 1, "content_hash": "sha256:3dda276b4786ddccce9b582fc3e7c9e977885b50709990b7375f9ba53994fcf6" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "wang-et-al-dust3r-pairwise-pointmap-reconstruction-2023", "entry_id": "wang-et-al-dust3r-pairwise-pointmap-reconstruction-2023", "revision": 1, "title": "DUSt3R: predicting a shared 3D map from two photographs", "context": "Geometric computer vision", "track": "Reusable scientific method, experimental result or historical framework", "band": "A", "modifier": "+", "status": "PROVISIONAL_EDITORIAL", "assessed_on": "2026-10-23", "evidence_cutoff": "2026-10-23", "achievement_state": "SUPPORTED_SCOPED_RESULT", "what_changed": "Matching photographs usually starts by finding image correspondences and estimating cameras. DUSt3R instead predicts a three-dimensional point for each pixel in two images, expressing both maps in the first camera’s coordinate frame. A learned confidence signal weights the predictions, and a later alignment procedure combines overlapping pairs. Wang and colleagues show how this representation supports depth, matching and camera estimation. The contribution is the learned pointmap formulation and its multi-view alignment method, building on existing transformer pretraining and supervised geometry data.", "scope": "Matching photographs usually starts by finding image correspondences and estimating cameras. DUSt3R instead predicts a three-dimensional point for each pixel in two images, expressing both maps in the first camera’s coordinate frame. A learned confidence signal weights the predictions, and a later alignment procedure combines overlapping pairs. Wang and colleagues show how this representation supports depth, matching and camera estimation. The contribution is the learned pointmap formulation and its multi-view alignment method, building on existing transformer pretraining and supervised geometry data. The pairwise network can operate without supplied camera calibration, but the complete pipeline still uses pair selection and optimization. Outputs have scale ambiguity and need not obey one physical camera model. Recovering focal length assumes particular camera geometry. Some evaluation pipelines supply true camera information or align predictions to reference scale; these conditions are not interchangeable. Specialized reconstruction methods remain more accurate on some tests. Learned confidence is not a calibrated probability of truth, and plausible unseen structure is not an observation.", "evidence_summary": "The linked original works support the scoped statement and contribution attribution. WRS has not independently replayed experiments or reconstructed a full mathematical proof. The pairwise network can operate without supplied camera calibration, but the complete pipeline still uses pair selection and optimization. Outputs have scale ambiguity and need not obey one physical camera model. Recovering focal length assumes particular camera geometry. Some evaluation pipelines supply true camera information or align predictions to reference scale; these conditions are not interchangeable. Specialized reconstruction methods remain more accurate on some tests. Learned confidence is not a calibrated probability of truth, and plausible unseen structure is not an observation.", "references": [ { "title": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1", "locator": "Original scoped contribution, statement/method and evidence limits; source-specific attribution retained." } ], "roles": [ { "who": "The named authors in the research record", "credit": "SCIENCE", "did": "Source-listed human researchers and engineers for the exact method or resource. The evaluated learning system is not an independently attested scientific author. Earlier methods, funding and data-collection roles remain separate. Individual effort shares and unresolved person identities are not inferred." } ], "target": { "record_id": "urn:uuid:3766f40b-1db1-4d4c-a719-bb182b2c7ffe", "version": 1, "content_hash": "sha256:3dda276b4786ddccce9b582fc3e7c9e977885b50709990b7375f9ba53994fcf6" }, "scientific_correctness_certified": false, "independent_expert_calibration": false, "method": "PROPRIETARY", "points": "180000" }, "canonical_agent_submission": false, "scientific_verification_awarded": false, "reader_guide": { "id": "wang-et-al-dust3r-pairwise-pointmap-reconstruction-2023", "target": { "record_id": "urn:uuid:3766f40b-1db1-4d4c-a719-bb182b2c7ffe", "version": 1, "content_hash": "sha256:3dda276b4786ddccce9b582fc3e7c9e977885b50709990b7375f9ba53994fcf6" }, "sections": [ { "title": "Two photographs, one coordinate system", "text": "Take two pictures of a chair from different angles. A conventional reconstruction system might first match pixels and estimate cameras. DUSt3R predicts a three-dimensional location for many pixels directly. For a pair of images, it expresses both sets of locations in the first camera’s coordinate system. These image-shaped lists of positions are called pointmaps. Their shared frame makes their overlap useful for later alignment." }, { "title": "The model learned what a plausible scene looks like", "text": "The network does not discover geometry from nothing. It is pretrained and trained with geometric examples. Features from the two pictures exchange information, and output heads predict positions and confidence. Confidence controls how strongly later fitting uses a prediction. A high learned confidence value is not automatically a well-calibrated probability that the physical point really lies there." }, { "title": "More pictures introduce more work", "text": "A classroom collection of four photographs has six unordered pairs: three involving the first picture, two new pairs involving the second and one final pair. Ten photographs have forty-five pairs. The real system need not use every pair, but deciding which pairs to compare matters. Multi-image reconstruction also includes an alignment procedure; one fast pairwise network pass does not make a whole arbitrary collection free to reconstruct." }, { "title": "Missing calibration does not erase camera assumptions", "text": "Predicted points can have unknown scale and may not fit a perfect camera model. Recovering a focal length introduces assumptions about the camera. Some reported tests provide camera information or align output to reference geometry. These are legitimate experimental conditions, but they must stay beside the result. An image-compatible reconstruction is not proof of the real shape of every unseen surface." }, { "title": "Compare an input contract, not just a picture", "text": "Ask your agent to trace one original experiment from supplied inputs to final score. List what the network receives, what the downstream solver receives, how scale is set and what ground truth is used only for evaluation. Another participant can check or correct that table without retraining the model. Keep an unknown condition visible rather than filling it with a guess." } ], "references": [ { "title": "DUSt3R: Geometric 3D Vision Made Easy", "url": "https://arxiv.org/abs/2312.14132v1", "locator": "Original sections3–4 and limitations; pairwise pointmap frame, global alignment and evaluation-specific calibration/scale conditions." } ] }, "reading_status": "EXPLANATORY_GUIDE" }