{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:M4ZW2724XH7HQFIQR4AE3MD7CZ","short_pith_number":"pith:M4ZW2724","canonical_record":{"source":{"id":"2302.12288","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-02-23T19:13:10Z","cross_cats_sorted":[],"title_canon_sha256":"2e5b8f8265ccb6ba846be9a73d9b5412d1953b6f90b81631fdef14f3d54c3bb7","abstract_canon_sha256":"722cbbb9276c09d8a5d6383e058b1d31f42e8500c36eb0624d9a760a0d8c0940"},"schema_version":"1.0"},"canonical_sha256":"67336d7f5cb9fe7815108f004db07f16491f1224e439febc40809f017c182b0c","source":{"kind":"arxiv","id":"2302.12288","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2302.12288","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"arxiv_version","alias_value":"2302.12288v1","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.12288","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"pith_short_12","alias_value":"M4ZW2724XH7H","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"M4ZW2724XH7HQFIQ","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"M4ZW2724","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:M4ZW2724XH7HQFIQR4AE3MD7CZ","target":"record","payload":{"canonical_record":{"source":{"id":"2302.12288","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-02-23T19:13:10Z","cross_cats_sorted":[],"title_canon_sha256":"2e5b8f8265ccb6ba846be9a73d9b5412d1953b6f90b81631fdef14f3d54c3bb7","abstract_canon_sha256":"722cbbb9276c09d8a5d6383e058b1d31f42e8500c36eb0624d9a760a0d8c0940"},"schema_version":"1.0"},"canonical_sha256":"67336d7f5cb9fe7815108f004db07f16491f1224e439febc40809f017c182b0c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:21.511283Z","signature_b64":"/7YbgJfoL9AQdH/I+VMDLHmx5p6qAgdoE9ZQeXGoKCC1jSyjxV/jwU0Q9fVCsqvHAaZVYiRznCMX8PTdGvJTDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"67336d7f5cb9fe7815108f004db07f16491f1224e439febc40809f017c182b0c","last_reissued_at":"2026-05-17T23:39:21.510617Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:21.510617Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2302.12288","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Vvg+eR7gMXNA4p+t98PiBfMM3wW/vLCPbw72Gb0tPcbdO0BA7jO3HVcnYxDg8gZQG/lhRwYRHlYeyZ++6/vEDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T04:10:18.413231Z"},"content_sha256":"a98c1f43d84010d003724bafc34fbe913616890c43285c5d75509ecd33970bb1","schema_version":"1.0","event_id":"sha256:a98c1f43d84010d003724bafc34fbe913616890c43285c5d75509ecd33970bb1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:M4ZW2724XH7HQFIQR4AE3MD7CZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ZoeDepth: Zero-shot Transfer by Combining Relative and Metric Depth","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A model pre-trained on relative depths from twelve datasets and fine-tuned on metric depth achieves strong zero-shot generalization while preserving scale.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Diana Wofk, Matthias M\\\"uller, Peter Wonka, Reiner Birkl, Shariq Farooq Bhat","submitted_at":"2023-02-23T19:13:10Z","abstract_excerpt":"This paper tackles the problem of depth estimation from a single image. Existing work either focuses on generalization performance disregarding metric scale, i.e. relative depth estimation, or state-of-the-art results on specific datasets, i.e. metric depth estimation. We propose the first approach that combines both worlds, leading to a model with excellent generalization performance while maintaining metric scale. Our flagship model, ZoeD-M12-NK, is pre-trained on 12 datasets using relative depth and fine-tuned on two datasets using metric depth. We use a lightweight head with a novel bin ad"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ZoeD-M12-NK is the first model that can jointly train on multiple datasets (NYU Depth v2 and KITTI) without a significant drop in performance and achieve unprecedented zero-shot generalization performance to eight unseen datasets from both indoor and outdoor domains.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The latent classifier reliably routes each image to the correct metric head without introducing errors that would degrade overall performance on mixed or ambiguous inputs.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ZoeDepth combines relative depth pre-training on many datasets with metric depth fine-tuning and automatic head routing to achieve strong zero-shot generalization while preserving metric scale.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A model pre-trained on relative depths from twelve datasets and fine-tuned on metric depth achieves strong zero-shot generalization while preserving scale.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"fe98e655a4f6af91568eb89b504908be85fc844013f135c743ea8893b22d7112"},"source":{"id":"2302.12288","kind":"arxiv","version":1},"verdict":{"id":"25e66e74-18cb-40ac-a55e-b819a3efe006","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T22:09:01.166711Z","strongest_claim":"ZoeD-M12-NK is the first model that can jointly train on multiple datasets (NYU Depth v2 and KITTI) without a significant drop in performance and achieve unprecedented zero-shot generalization performance to eight unseen datasets from both indoor and outdoor domains.","one_line_summary":"ZoeDepth combines relative depth pre-training on many datasets with metric depth fine-tuning and automatic head routing to achieve strong zero-shot generalization while preserving metric scale.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The latent classifier reliably routes each image to the correct metric head without introducing errors that would degrade overall performance on mixed or ambiguous inputs.","pith_extraction_headline":"A model pre-trained on relative depths from twelve datasets and fine-tuned on metric depth achieves strong zero-shot generalization while preserving scale."},"references":{"count":52,"sample":[{"doi":"","year":1972,"title":"Miton Abramowitz. Stegun., ia (1972). handbook of mathe- matical functions. Formulas, Graphs and Mathematical Ta- bles, 2002. 4","work_id":"bf9098b2-e0f4-4d34-bd70-11464cc82783","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Attention attention everywhere: Monocular depth prediction with skip attention","work_id":"74105095-d6d7-4366-ace0-2e5be48bc7ac","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"BEiT: BERT Pre-Training of Image Transformers","work_id":"d74eda3c-bf7e-45f1-a8f1-a0137ecca3f4","ref_index":3,"cited_arxiv_id":"2106.08254","is_internal_anchor":true},{"doi":"","year":null,"title":"Unimodal prob- ability distributions for deep ordinal classification","work_id":"f66b3823-1fbc-42d2-aa51-ac55ee013ea2","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2017,"title":"PMLR, 06–11 Aug 2017. 4, 9","work_id":"6e80c81d-d687-46d1-b85b-1031126f4f37","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":52,"snapshot_sha256":"43de24ab6e9d8402af7154929f09f0f7578632d138087dfe5091c3fd9d659984","internal_anchors":1},"formal_canon":{"evidence_count":2,"snapshot_sha256":"1e6c33a4e5f6d7db5f3e50525ade333bdcc72416de32bc9483c73676340531c3"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"25e66e74-18cb-40ac-a55e-b819a3efe006"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eHQ5tIoKl0OWF057hlJANawP6nQz2oHWom7ZHFsjvJvliqQ46Sx12umGHqMgNmAxCzT18smqdi9819eTIGWVAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T04:10:18.414481Z"},"content_sha256":"390d5c565af63f3337c9edf1e87df6ee33f63ef961b297c18ef50c471be0129e","schema_version":"1.0","event_id":"sha256:390d5c565af63f3337c9edf1e87df6ee33f63ef961b297c18ef50c471be0129e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/M4ZW2724XH7HQFIQR4AE3MD7CZ/bundle.json","state_url":"https://pith.science/pith/M4ZW2724XH7HQFIQR4AE3MD7CZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/M4ZW2724XH7HQFIQR4AE3MD7CZ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-26T04:10:18Z","links":{"resolver":"https://pith.science/pith/M4ZW2724XH7HQFIQR4AE3MD7CZ","bundle":"https://pith.science/pith/M4ZW2724XH7HQFIQR4AE3MD7CZ/bundle.json","state":"https://pith.science/pith/M4ZW2724XH7HQFIQR4AE3MD7CZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/M4ZW2724XH7HQFIQR4AE3MD7CZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:M4ZW2724XH7HQFIQR4AE3MD7CZ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"722cbbb9276c09d8a5d6383e058b1d31f42e8500c36eb0624d9a760a0d8c0940","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-02-23T19:13:10Z","title_canon_sha256":"2e5b8f8265ccb6ba846be9a73d9b5412d1953b6f90b81631fdef14f3d54c3bb7"},"schema_version":"1.0","source":{"id":"2302.12288","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2302.12288","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"arxiv_version","alias_value":"2302.12288v1","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.12288","created_at":"2026-05-17T23:39:21Z"},{"alias_kind":"pith_short_12","alias_value":"M4ZW2724XH7H","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"M4ZW2724XH7HQFIQ","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"M4ZW2724","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:390d5c565af63f3337c9edf1e87df6ee33f63ef961b297c18ef50c471be0129e","target":"graph","created_at":"2026-05-17T23:39:21Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"ZoeD-M12-NK is the first model that can jointly train on multiple datasets (NYU Depth v2 and KITTI) without a significant drop in performance and achieve unprecedented zero-shot generalization performance to eight unseen datasets from both indoor and outdoor domains."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The latent classifier reliably routes each image to the correct metric head without introducing errors that would degrade overall performance on mixed or ambiguous inputs."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"ZoeDepth combines relative depth pre-training on many datasets with metric depth fine-tuning and automatic head routing to achieve strong zero-shot generalization while preserving metric scale."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A model pre-trained on relative depths from twelve datasets and fine-tuned on metric depth achieves strong zero-shot generalization while preserving scale."}],"snapshot_sha256":"fe98e655a4f6af91568eb89b504908be85fc844013f135c743ea8893b22d7112"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"1e6c33a4e5f6d7db5f3e50525ade333bdcc72416de32bc9483c73676340531c3"},"paper":{"abstract_excerpt":"This paper tackles the problem of depth estimation from a single image. Existing work either focuses on generalization performance disregarding metric scale, i.e. relative depth estimation, or state-of-the-art results on specific datasets, i.e. metric depth estimation. We propose the first approach that combines both worlds, leading to a model with excellent generalization performance while maintaining metric scale. Our flagship model, ZoeD-M12-NK, is pre-trained on 12 datasets using relative depth and fine-tuned on two datasets using metric depth. We use a lightweight head with a novel bin ad","authors_text":"Diana Wofk, Matthias M\\\"uller, Peter Wonka, Reiner Birkl, Shariq Farooq Bhat","cross_cats":[],"headline":"A model pre-trained on relative depths from twelve datasets and fine-tuned on metric depth achieves strong zero-shot generalization while preserving scale.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-02-23T19:13:10Z","title":"ZoeDepth: Zero-shot Transfer by Combining Relative and Metric Depth"},"references":{"count":52,"internal_anchors":1,"resolved_work":52,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Miton Abramowitz. Stegun., ia (1972). handbook of mathe- matical functions. Formulas, Graphs and Mathematical Ta- bles, 2002. 4","work_id":"bf9098b2-e0f4-4d34-bd70-11464cc82783","year":1972},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Attention attention everywhere: Monocular depth prediction with skip attention","work_id":"74105095-d6d7-4366-ace0-2e5be48bc7ac","year":2022},{"cited_arxiv_id":"2106.08254","doi":"","is_internal_anchor":true,"ref_index":3,"title":"BEiT: BERT Pre-Training of Image Transformers","work_id":"d74eda3c-bf7e-45f1-a8f1-a0137ecca3f4","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Unimodal prob- ability distributions for deep ordinal classification","work_id":"f66b3823-1fbc-42d2-aa51-ac55ee013ea2","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"PMLR, 06–11 Aug 2017. 4, 9","work_id":"6e80c81d-d687-46d1-b85b-1031126f4f37","year":2017}],"snapshot_sha256":"43de24ab6e9d8402af7154929f09f0f7578632d138087dfe5091c3fd9d659984"},"source":{"id":"2302.12288","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T22:09:01.166711Z","id":"25e66e74-18cb-40ac-a55e-b819a3efe006","model_set":{"reader":"grok-4.3"},"one_line_summary":"ZoeDepth combines relative depth pre-training on many datasets with metric depth fine-tuning and automatic head routing to achieve strong zero-shot generalization while preserving metric scale.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A model pre-trained on relative depths from twelve datasets and fine-tuned on metric depth achieves strong zero-shot generalization while preserving scale.","strongest_claim":"ZoeD-M12-NK is the first model that can jointly train on multiple datasets (NYU Depth v2 and KITTI) without a significant drop in performance and achieve unprecedented zero-shot generalization performance to eight unseen datasets from both indoor and outdoor domains.","weakest_assumption":"The latent classifier reliably routes each image to the correct metric head without introducing errors that would degrade overall performance on mixed or ambiguous inputs."}},"verdict_id":"25e66e74-18cb-40ac-a55e-b819a3efe006"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:a98c1f43d84010d003724bafc34fbe913616890c43285c5d75509ecd33970bb1","target":"record","created_at":"2026-05-17T23:39:21Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"722cbbb9276c09d8a5d6383e058b1d31f42e8500c36eb0624d9a760a0d8c0940","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-02-23T19:13:10Z","title_canon_sha256":"2e5b8f8265ccb6ba846be9a73d9b5412d1953b6f90b81631fdef14f3d54c3bb7"},"schema_version":"1.0","source":{"id":"2302.12288","kind":"arxiv","version":1}},"canonical_sha256":"67336d7f5cb9fe7815108f004db07f16491f1224e439febc40809f017c182b0c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"67336d7f5cb9fe7815108f004db07f16491f1224e439febc40809f017c182b0c","first_computed_at":"2026-05-17T23:39:21.510617Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:21.510617Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"/7YbgJfoL9AQdH/I+VMDLHmx5p6qAgdoE9ZQeXGoKCC1jSyjxV/jwU0Q9fVCsqvHAaZVYiRznCMX8PTdGvJTDA==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:21.511283Z","signed_message":"canonical_sha256_bytes"},"source_id":"2302.12288","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a98c1f43d84010d003724bafc34fbe913616890c43285c5d75509ecd33970bb1","sha256:390d5c565af63f3337c9edf1e87df6ee33f63ef961b297c18ef50c471be0129e"],"state_sha256":"8ab8603458ee0159df0614b12ec126d959081e6a38e4b5eb7d620c53cd841687"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rjplI7Li50pl71DzH/Vc/xUgs5G8KjUO/Q9EONlr+7bSKYmqCip/TEKa2r6bwAVlMF6SsKb7J00JqpZbttZQAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T04:10:18.419181Z","bundle_sha256":"857944397f4f0ac495a6b72a87eb819214e3a805963eed69175bbc46d8ee2cca"}}