{"paper":{"title":"Benchmarking Audio Deepfake Detection Robustness in Real-world Communication Scenarios","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A data augmentation strategy improves audio deepfake detection performance under codec compression and packet loss.","cross_cats":["eess.SP"],"primary_cat":"eess.AS","authors_text":"Haohan Shi, Safak Dogan, Saif Alzubi, Tianjin Huang, Xiyu Shi, Yunxiao Zhang","submitted_at":"2025-04-16T18:44:05Z","abstract_excerpt":"Existing Audio Deepfake Detection (ADD) systems often struggle to generalise effectively due to the significantly degraded audio quality caused by audio codec compression and channel transmission effects in real-world communication scenarios. To address this challenge, we developed a rigorous benchmark to evaluate the performance of the ADD system under such scenarios. We introduced ADD-C, a new test dataset to evaluate the robustness of ADD systems under diverse communication conditions, including different combinations of audio codecs for compression and packet loss rates. Benchmarking three"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experimental results demonstrated that the proposed approach significantly enhances the performance of ADD systems on the proposed ADD-C dataset.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The combinations of audio codecs and packet loss rates used to build the ADD-C dataset accurately represent the degradations that occur in actual real-world communication channels.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"The paper creates ADD-C benchmark dataset for audio deepfake detection under codec compression and packet loss, shows baseline degradation, and demonstrates a data augmentation method that boosts robustness.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A data augmentation strategy improves audio deepfake detection performance under codec compression and packet loss.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7bb9e26754013510407b5fb85850033de3a7734c756f174fda7ae8812d248f36"},"source":{"id":"2504.12423","kind":"arxiv","version":4},"verdict":{"id":"ccc6f7b0-9fee-4670-8d82-d63e988b2b9b","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-22T19:40:23.449731Z","strongest_claim":"Experimental results demonstrated that the proposed approach significantly enhances the performance of ADD systems on the proposed ADD-C dataset.","one_line_summary":"The paper creates ADD-C benchmark dataset for audio deepfake detection under codec compression and packet loss, shows baseline degradation, and demonstrates a data augmentation method that boosts robustness.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The combinations of audio codecs and packet loss rates used to build the ADD-C dataset accurately represent the degradations that occur in actual real-world communication channels.","pith_extraction_headline":"A data augmentation strategy improves audio deepfake detection performance under codec compression and packet loss."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2504.12423/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":2,"snapshot_sha256":"dbd53bd0eda65db3c008f51ff37f9d97814d79492bd9fd348c7e538fb910f9d6"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}