{"paper":{"title":"TVRN: Invertible Neural Networks for Compression-Aware Temporal Video Rescaling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"TVRN's invertible architecture and surrogate network enable end-to-end compression-aware video frame rate rescaling.","cross_cats":["cs.CV"],"primary_cat":"eess.IV","authors_text":"Dong Liu, Feng Wu, Li Li, Xinmin Feng","submitted_at":"2026-05-15T03:44:14Z","abstract_excerpt":"To fit diverse display and bandwidth constraints, high-frame-rate videos are temporally downscaled to low-frame-rate (LFR) and later upscaled, requiring joint optimization for effective frame-rate rescaling. However, existing methods typically link the two operations via training objectives, without fully exploiting their reciprocal nature, which may cause high-frequency information loss. Moreover, they overlook the impact of lossy codecs on LFR videos, limiting real-world applicability. In this work, we propose an end-to-end framework for compression-aware frame-rate rescaling, named TVRN. To"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Extensive experiments show that TVRN outperforms existing methods in reconstruction quality under industrial video compression settings.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The surrogate network provides sufficiently accurate gradient approximations for non-differentiable lossy codecs so that end-to-end optimization remains stable and does not introduce artifacts that would not appear with the true codec.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"TVRN combines invertible wavelet-based networks with a surrogate gradient approximator and compression-aware asymmetric design to improve frame-rate rescaling quality under real codecs.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"TVRN's invertible architecture and surrogate network enable end-to-end compression-aware video frame rate rescaling.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"d8b5d7fa807b62f241cf344ef592603b1e38033222ddfc829fd7112bf87ef911"},"source":{"id":"2605.15579","kind":"arxiv","version":1},"verdict":{"id":"989ca850-7b3f-4ddb-9fbb-7e201a8e74ad","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T19:49:11.860381Z","strongest_claim":"Extensive experiments show that TVRN outperforms existing methods in reconstruction quality under industrial video compression settings.","one_line_summary":"TVRN combines invertible wavelet-based networks with a surrogate gradient approximator and compression-aware asymmetric design to improve frame-rate rescaling quality under real codecs.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The surrogate network provides sufficiently accurate gradient approximations for non-differentiable lossy codecs so that end-to-end optimization remains stable and does not introduce artifacts that would not appear with the true codec.","pith_extraction_headline":"TVRN's invertible architecture and surrogate network enable end-to-end compression-aware video frame rate rescaling."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.15579/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_compliance","ran_at":"2026-05-19T20:01:43.264213Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T20:01:19.293391Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T19:34:35.248938Z","status":"skipped","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T17:41:56.071384Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"03c83bedbb01879c69b0694bbd96a18f944496310f40bc9950fd92091995bded"},"references":{"count":89,"sample":[{"doi":"","year":2019,"title":"BETA: bandwidth-efficient temporal adaptation for video streaming over reliable transports,","work_id":"f56271e9-729f-4d6c-9c63-bdc7efdf2c89","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"VOXEL: Cross-layer optimization for video streaming with imperfect transmission,","work_id":"94f5c4a5-60d6-4b9e-a280-c2992a2b1405","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Reparo: Qoe-aware live video streaming in low- rate networks by intelligent frame recovery,","work_id":"228c677f-69fa-44d7-90db-7e978832b3f5","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Enabling high quality Real-Time communications with adaptive Frame-Rate,","work_id":"830a29a5-5595-4c59-8cc8-240f49917399","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"SAFR: A real- time communication system with adaptive frame rate,","work_id":"207d4cca-bdc4-4a6d-b054-fd198093e961","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":89,"snapshot_sha256":"423373122817901a498489ef73349040cee26679c9942fbc1e7ad638f2f4ac56","internal_anchors":5},"formal_canon":{"evidence_count":2,"snapshot_sha256":"d643da7ab9e47265d738d1c9357e7309a914f4f81acf6ce321f7a8531772277a"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}