{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:5LCVXXSPIUDEM6BQWNSSTRKA5T","short_pith_number":"pith:5LCVXXSP","schema_version":"1.0","canonical_sha256":"eac55bde4f4506467830b36529c540ecdc89b43c6f6c0d678c1085a531b7a2f3","source":{"kind":"arxiv","id":"2012.14950","version":2},"attestation_state":"computed","paper":{"title":"2D or not 2D? Adaptive 3D Convolution Selection for Efficient Video Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Abhinav Shrivastava, Hengduo Li, Larry S. Davis, Zuxuan Wu","submitted_at":"2020-12-29T21:40:38Z","abstract_excerpt":"3D convolutional networks are prevalent for video recognition. While achieving excellent recognition performance on standard benchmarks, they operate on a sequence of frames with 3D convolutions and thus are computationally demanding. Exploiting large variations among different videos, we introduce Ada3D, a conditional computation framework that learns instance-specific 3D usage policies to determine frames and convolution layers to be used in a 3D network. These policies are derived with a two-head lightweight selection network conditioned on each input video clip. Then, only frames and convo"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2012.14950","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2020-12-29T21:40:38Z","cross_cats_sorted":[],"title_canon_sha256":"f2cf1e380ad25dd80db88dded90fab84ad04bee4ad5a6e6cde8feea43198391a","abstract_canon_sha256":"5b14b9b5727fd976859bc72cb9bb6bf85c151cef4b68ed1bd5182df20131152c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:36:09.472354Z","signature_b64":"snEu/LH6b8mx5P8qQ6TH3OATLfpmiP6aXCIh2NOFPtNbx01uEiIguqeXQvsNuSsTpcG2NzyWkCEGUOPbR73hAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eac55bde4f4506467830b36529c540ecdc89b43c6f6c0d678c1085a531b7a2f3","last_reissued_at":"2026-07-05T02:36:09.471904Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:36:09.471904Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"2D or not 2D? Adaptive 3D Convolution Selection for Efficient Video Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Abhinav Shrivastava, Hengduo Li, Larry S. Davis, Zuxuan Wu","submitted_at":"2020-12-29T21:40:38Z","abstract_excerpt":"3D convolutional networks are prevalent for video recognition. While achieving excellent recognition performance on standard benchmarks, they operate on a sequence of frames with 3D convolutions and thus are computationally demanding. Exploiting large variations among different videos, we introduce Ada3D, a conditional computation framework that learns instance-specific 3D usage policies to determine frames and convolution layers to be used in a 3D network. These policies are derived with a two-head lightweight selection network conditioned on each input video clip. Then, only frames and convo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2012.14950","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2012.14950/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2012.14950","created_at":"2026-07-05T02:36:09.471967+00:00"},{"alias_kind":"arxiv_version","alias_value":"2012.14950v2","created_at":"2026-07-05T02:36:09.471967+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2012.14950","created_at":"2026-07-05T02:36:09.471967+00:00"},{"alias_kind":"pith_short_12","alias_value":"5LCVXXSPIUDE","created_at":"2026-07-05T02:36:09.471967+00:00"},{"alias_kind":"pith_short_16","alias_value":"5LCVXXSPIUDEM6BQ","created_at":"2026-07-05T02:36:09.471967+00:00"},{"alias_kind":"pith_short_8","alias_value":"5LCVXXSP","created_at":"2026-07-05T02:36:09.471967+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T","json":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T.json","graph_json":"https://pith.science/api/pith-number/5LCVXXSPIUDEM6BQWNSSTRKA5T/graph.json","events_json":"https://pith.science/api/pith-number/5LCVXXSPIUDEM6BQWNSSTRKA5T/events.json","paper":"https://pith.science/paper/5LCVXXSP"},"agent_actions":{"view_html":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T","download_json":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T.json","view_paper":"https://pith.science/paper/5LCVXXSP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2012.14950&json=true","fetch_graph":"https://pith.science/api/pith-number/5LCVXXSPIUDEM6BQWNSSTRKA5T/graph.json","fetch_events":"https://pith.science/api/pith-number/5LCVXXSPIUDEM6BQWNSSTRKA5T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T/action/storage_attestation","attest_author":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T/action/author_attestation","sign_citation":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T/action/citation_signature","submit_replication":"https://pith.science/pith/5LCVXXSPIUDEM6BQWNSSTRKA5T/action/replication_record"}},"created_at":"2026-07-05T02:36:09.471967+00:00","updated_at":"2026-07-05T02:36:09.471967+00:00"}