{"as_of":"2026-08-09T01:02:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:427372ac3610c21dbb4a57ca557e6f881e69f10704a9342a2c191a08150a4815","coverage":[{"denominator":44,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":44,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-01T22:35:52.025790Z","state":"measured"},{"denominator":44,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":44,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-08T06:32:00.761636+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2607.15697/citation-record","integrity":"/paper/2607.15697/integrity","json":"/paper/2607.15697/citation-record.json","paper":"/paper/2607.15697"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:45.735201Z","title":"[9] first proposed theBad- Netsattack scheme based on data poisoning for outsourced training and transfer learning scenarios","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:45.735201Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:769f19299a16805f2e6c1b0f28f0b766e6e6e9e1caf6492631d15344b39bb2b4","observation_id":"50f123c8-fd16-47fd-b516-46f23940ebd6","resolution":{"observed_at":"2026-08-01T22:35:45.735201Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:45.811467Z","title":"Liu et al","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:45.811467Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:e7c8a8a78a8cad09a154318de601344c9b275eb3f52bf238dadbb9c089e976df","observation_id":"7d47f12e-dfc6-4fcc-a03b-33aef6cc9b14","resolution":{"observed_at":"2026-08-01T22:35:45.811467Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:46.169554Z","title":"Koffas et al","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:46.169554Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:5f0efce5e1fd21140c2ecca50308e44e592489afcac5c6199046dfbe9d867cc4","observation_id":"921c1795-c393-4bce-ab3f-56f3dd95401a","resolution":{"observed_at":"2026-08-01T22:35:46.169554Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:49.419113Z","title":"Can you hear it? backdoor attacks via ultrasonic triggers","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:49.419113Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:db61dde3abd0dd8cb4e0a122e0443f50e2892830db3f50b9cee79689810736db","observation_id":"edb41954-4598-4c63-8db2-fb31436c28e0","resolution":{"observed_at":"2026-08-01T22:35:49.419113Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:46.323715Z","title":"For a speech command recognition task, the objective is to learn a modelF θ :X→Y, whereXdenotes the input space andYdenotes the label space","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:46.323715Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:d3eb1696743193f62ba3978766f7e4848aabb6bbc9136d4d7e9aa8f11cb67bfc","observation_id":"ac61d621-54ff-490a-9a5e-e960263508c8","resolution":{"observed_at":"2026-08-01T22:35:46.323715Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:46.470074Z","title":"Followed by incorporating the poisoned samples to create the poisoning training set:D ∗ train =D train S Dpoison","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:46.470074Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:1be83735b5f9d97eb9a52284633d070386ad44ed8bab75e579205ac3ec399853","observation_id":"73ed4ec8-3238-4e1b-907c-6ec5965b26ab","resolution":{"observed_at":"2026-08-01T22:35:46.470074Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:46.669622Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:46.669622Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:efc0cb0301cdd431f22aac56a328752e078a6f1c7634ca194ceceef99d2fc424","observation_id":"72f24ae4-2b4e-476e-8ca2-1e5f08f609a8","resolution":{"observed_at":"2026-08-01T22:35:46.669622Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:46.831932Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:46.831932Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:3b852c5f87adbb6887db5e1935869aaf24d503aaf3758bcf34a6e9aec662610d","observation_id":"f284866d-d1db-4c76-bf3f-3e6ad29fc337","resolution":{"observed_at":"2026-08-01T22:35:46.831932Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:47.002250Z","title":"In contrast, clean samples undergoing the same robust perturbations exhibit substantial alterations in the prediction results","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:47.002250Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:2d1e24f445cdf4306e1b3535efea9d7cdf97aa56e214d1daf6786885aff20971","observation_id":"73a62bba-d356-4a22-91f1-8c8715fa6b7a","resolution":{"observed_at":"2026-08-01T22:35:47.002250Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:47.145183Z","title":"These per- turbations are derived either from the dataset or random noise","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:47.145183Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:cd41d2840a6cce505b991feb37c6fb5e88d85b2563b4616f6ec1a8f935b91004","observation_id":"3843d623-2f10-4501-9b8e-ce885bc41ef1","resolution":{"observed_at":"2026-08-01T22:35:47.145183Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:47.284600Z","title":null,"venue":null,"work_id":null,"year":2048},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:47.284600Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:1f51ca791621a15e7827a58f8c4895e0222e642c7aa5669b7670bae5b5a426df","observation_id":"4bef5cba-76e8-4c7e-b1bc-4ba7030ea5d4","resolution":{"observed_at":"2026-08-01T22:35:47.284600Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:47.437387Z","title":"In this study, we propose an autoencoder architecture (Fig","venue":null,"work_id":null,"year":2048},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:47.437387Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:19911ce978f1a0053cbad7660326cf4ad9dad52dfad4faebb69d675b178c7daf","observation_id":"970e453b-67c8-4ab1-94d8-91eb31656ee5","resolution":{"observed_at":"2026-08-01T22:35:47.437387Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:47.598071Z","title":"Below is a brief description: •SCDv2: Speech Commands Dataset Version 2 (SCDv2)","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:47.598071Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:2f1bbe66a77247884540d68bac02b0205c56bb7825a738890a2be0d3047945ba","observation_id":"93509a30-a0cd-4af1-bf14-169fcf3af5b0","resolution":{"observed_at":"2026-08-01T22:35:47.598071Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:50.874500Z","title":"Fine-pruning: Defending against backdooring attacks on deep neural networks","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:50.874500Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:4c469563fbb244cb02784d2212e420edf6822fff2af8126177dae0af3277c098","observation_id":"5200458d-c68a-4012-82e3-178615cae3b4","resolution":{"observed_at":"2026-08-01T22:35:50.874500Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:47.868284Z","title":"The triggers can be categorized into three types, each illus- trated in the spectrogram depicted in Fig","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:47.868284Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:ed2863af7418ead45a4af93e1e60107155686aa6be8a2b1ee7488ef5b003d538","observation_id":"67647c67-48c5-4ff6-9fc2-c3b963572998","resolution":{"observed_at":"2026-08-01T22:35:47.868284Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:47.996294Z","title":"•Attack Success Rate (ASR): This metric quantifies the proportion of poisoned samples successfully directed the target label","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:47.996294Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:2b9b044434427c06878041be397c763771476df85dfca26c31435cefd6b1a614","observation_id":"39d8166a-66fd-44c4-bf32-b4cffd45cad8","resolution":{"observed_at":"2026-08-01T22:35:47.996294Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:45.994746Z","title":"Gao et al","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:45.994746Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:d864b1eb5192aee2ce635aefa8a967f2ec512b3470c99c23bba7032418f968ac","observation_id":"59f0e8d0-4c8e-4454-8cae-455534749b48","resolution":{"observed_at":"2026-08-01T22:35:45.994746Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:48.124564Z","title":"Table I presents the attack performance of the backdoor model","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:48.124564Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:996d24d76f60a222e237ffc1763852e7d0da09043508da743d747ebb05a62558","observation_id":"0e33ce52-7cdd-413c-b804-0a48d9ec8bdd","resolution":{"observed_at":"2026-08-01T22:35:48.124564Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:47.731632Z","title":"We have selected 10 commands to form a 10-class speech recognition task","venue":null,"work_id":null,"year":2048},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:47.731632Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:8ef31a5aa8b25125cd914e81602b24178d1c5786eec3c24f57114476884d9a0f","observation_id":"4f742287-f9fa-497d-a5af-76e7f7c839fb","resolution":{"observed_at":"2026-08-01T22:35:47.731632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:48.311458Z","title":"We set SNR to 10 and determine the mixing ratio using (5)","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:48.311458Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:74ef1a5de6167017387749de962197a778eda000059f1967ae05271499358782","observation_id":"8da0e73e-9fb6-407d-8684-f67cdcbb7b17","resolution":{"observed_at":"2026-08-01T22:35:48.311458Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:48.471193Z","title":"In speech recognition tasks, S-STRIP demonstrates effective defense capabilities","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:48.471193Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:19100df420f8f0a144b86943ae1e073c587d21e23c5080454fd02a9b78cb1b56","observation_id":"e898f1c2-d415-4478-ab9e-03ec0296d45e","resolution":{"observed_at":"2026-08-01T22:35:48.471193Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:48.615248Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:48.615248Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:c2b86e4ad8a8f5755ee1bb6c4c71d4443b2e2739b1b6e7e2da9d63a9fdd8096a","observation_id":"8be256ad-9b39-4864-9d52-6f22fb5e6ec9","resolution":{"observed_at":"2026-08-01T22:35:48.615248Z","resolver_source":null,"status":"malformed_identifier"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:48.743803Z","title":"Following purification by the autoencoder, the threat posed by the poisoned samples is significantly mitigated, with ASR decreasing by more than 90% across all cases","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:48.743803Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:d4c87baa0db7f9f608a0087ceb48f65c0e1eeb783e5bb1200480a04f8e5190f2","observation_id":"b5749738-dae4-4522-9993-849691bb0ace","resolution":{"observed_at":"2026-08-01T22:35:48.743803Z","resolver_source":null,"status":"malformed_identifier"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:48.888627Z","title":"Adaptive square attack: Fooling autonomous cars with adversarial traffic signs","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:48.888627Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:e21457b852295058c47518c3d30b3a3b70400a20b424ca35dedb2372d942c091","observation_id":"e740a509-805e-42d2-ba04-9fc2441444ca","resolution":{"observed_at":"2026-08-01T22:35:48.888627Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:49.085246Z","title":"Data poisoning attacks to deep learning based recom- mender systems","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:49.085246Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:b8d1df9dce5d801aacfaec1cb144b59b26064cc431e52e2c262efe3d7414752c","observation_id":"7da2afa0-fd97-4b1a-8ea1-2e5a10e2ed59","resolution":{"observed_at":"2026-08-01T22:35:49.085246Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:49.244588Z","title":"Backdoor learning: A survey.IEEE Transactions on Neural Networks and Learning Systems, pages 1–18, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:49.244588Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:a15325953a625fd87061beca6504303fc9788bba98bff0fb908f020bc3dfb724","observation_id":"572f2ccd-b6c8-4d7c-8ad7-6f17bf0922fa","resolution":{"observed_at":"2026-08-01T22:35:49.244588Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:49.590200Z","title":"Oppor- tunistic backdoor attacks: Exploring human-imperceptible vulnerabilities on speech recognition systems","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:49.590200Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:ee08838d3416df405a4fb0e495d16d6ac4205192eb937c48a60c3cebea24f45c","observation_id":"ce4c05e0-0034-451e-9a02-7f9dc4ae89ac","resolution":{"observed_at":"2026-08-01T22:35:49.590200Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:49.746735Z","title":"Trojanmodel: A practical trojan attack against automatic speech recognition systems","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:49.746735Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:abdafd74206a8dcbf797ceadac61dea3248177c1f3c9f0a4017c447afc20e318","observation_id":"63fd2a4c-5f75-48ee-a3e3-cd07dbb07cac","resolution":{"observed_at":"2026-08-01T22:35:49.746735Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:49.887640Z","title":"Strip: A defence against trojan attacks on deep neural networks","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:49.887640Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:6cf6c8ffd2af91ec799cb27da4ad6acd4f4a49327e70f25677ee2eeda47615bd","observation_id":"7cc55a93-0c16-4854-bab9-364bd9358868","resolution":{"observed_at":"2026-08-01T22:35:49.887640Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:50.011265Z","title":"Anti-backdoor learning: Training clean models on poisoned data","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:50.011265Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:bb933b3a0ec83777895418b82572c08950df9077a81fd90716d5414f1a9cd287","observation_id":"d1731557-473e-4361-80f0-219254eff31f","resolution":{"observed_at":"2026-08-01T22:35:50.011265Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:50.134123Z","title":"Badnets: Evaluating backdooring attacks on deep neural networks.IEEE Access, 7:47230–47244, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:50.134123Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:41fee962137636a7264fe9fc0a6342c8d12987c6b485fbddf803e805a46774c5","observation_id":"6702d77f-ff98-4a72-a6d7-897a819056d3","resolution":{"observed_at":"2026-08-01T22:35:50.134123Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:50.257640Z","title":"Trojaning attack on neural networks","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:50.257640Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:cda7f83d62a570f8e9b5c5a41d2d9f9a1c4a9ba934ce39dc8f14333c50f72976","observation_id":"aa3f13c6-3047-47c3-bbcc-55f862462c38","resolution":{"observed_at":"2026-08-01T22:35:50.257640Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1712.05526","last_updated":"2017-12-15T04:26:26Z","snapshot_observed_at":"2026-07-06T06:14:30.795326Z","submitted_at":"2017-12-15T04:26:26Z","title":"Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1712.05526","snapshot_observed_at":"2026-08-01T22:35:50.423148Z","title":"Tar- geted backdoor attacks on deep learning systems using data poisoning","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:50.423148Z"},"links":{"cited_paper":"/paper/1712.05526","citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:89f9a4d24aa9bbec086325f8dd691bd75d89020e38dbdd2347f03d60ac67805c","observation_id":"b0240ce9-20a0-4bb0-af90-a187cfaaf5fd","resolution":{"observed_at":"2026-08-01T22:35:50.423148Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:50.567354Z","title":"Reflection backdoor: A natural backdoor attack on deep neural networks","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:50.567354Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:18a8bb4d057bf9072cb7e4f0258d9046282dccba59cdb2a9bf8c041b62ed1517","observation_id":"8fbecb7c-e449-4fb7-8f84-c708273e628a","resolution":{"observed_at":"2026-08-01T22:35:50.567354Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:50.689165Z","title":"Deep feature space trojan attack of neural networks by controlled detoxifica- tion","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:50.689165Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:e1ac37e42bfc540826d9fc096adbf7d3f6baf8c1aafcbcc478a64a3c54bb78ef","observation_id":"50dba290-92f2-48b5-9817-41387186a0ab","resolution":{"observed_at":"2026-08-01T22:35:50.689165Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:51.005408Z","title":"Neural cleanse: Identifying and miti- gating backdoor attacks in neural networks","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:51.005408Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:0253033c34a4a5daae1765bcce75f7019769861b9dc03a873fbe92275b76bf83","observation_id":"b5bcd29f-3317-4152-8236-79fba68168bf","resolution":{"observed_at":"2026-08-01T22:35:51.005408Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:51.123950Z","title":"Februus: Input purification defense against trojan attacks on deep neural network systems","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:51.123950Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:fbd248bd1796230e0b9abb70a0eccc676bfe02cbaeba2052020317e0c6e9f4a9","observation_id":"38ba3416-a096-442d-ba64-7ca89da755a7","resolution":{"observed_at":"2026-08-01T22:35:51.123950Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:51.281168Z","title":"Grad-cam: Visual explanations from deep networks via gradient-based localization","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:51.281168Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:deb7aaac11f4afbcb667b41a1611558916cf53185ccff120f75251001b13f640","observation_id":"f4fcbdd2-70bb-429c-b593-512181de7d0f","resolution":{"observed_at":"2026-08-01T22:35:51.281168Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:51.408809Z","title":"Backdoor attack against speaker verification","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:51.408809Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:ba02b6bedf75f93cacfe0be23cf429b5f389135d9bc93d72c9e2be4708e91cd0","observation_id":"6a3571f6-ca6b-455c-a418-df682e87713f","resolution":{"observed_at":"2026-08-01T22:35:51.408809Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1804.03209","last_updated":"2018-04-09T19:58:17Z","snapshot_observed_at":"2026-07-06T06:32:32.083176Z","submitted_at":"2018-04-09T19:58:17Z","title":"Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1804.03209","snapshot_observed_at":"2026-08-01T22:35:51.525874Z","title":"Speech commands: A dataset for limited-vocabulary speech recognition.arXiv preprint arXiv:1804.03209, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:51.525874Z"},"links":{"cited_paper":"/paper/1804.03209","citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:3260e372c9256e132b435e5a172197fd79d39a8b433840b51e522b50ffbb399e","observation_id":"e565572e-067c-44b0-afed-ab3e4a2ca778","resolution":{"observed_at":"2026-08-01T22:35:51.525874Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1807.03418","last_updated":"2023-11-27T18:26:32Z","snapshot_observed_at":"2026-08-06T07:51:43.535794Z","submitted_at":"2018-07-09T23:11:17Z","title":"AudioMNIST: Exploring Explainable Artificial Intelligence for Audio Analysis on a Simple Benchmark","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1807.03418","snapshot_observed_at":"2026-08-01T22:35:51.673729Z","title":"Interpreting and explaining deep neural networks for classification of audio signals.arXiv preprint arXiv:1807.03418, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:51.673729Z"},"links":{"cited_paper":"/paper/1807.03418","citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:c7d84d9e78b666be030a9654d5df17eb875a791fe88521eb79fe5de78ee40a7d","observation_id":"c4603685-7537-48a5-8ac5-8b7be17d6106","resolution":{"observed_at":"2026-08-01T22:35:51.673729Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:51.781237Z","title":"Ad- versarial example detection by classification for deep speech recognition","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:51.781237Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:a91a0e2a4259974019b6a8c86b2ec660d09ebfefaa70a57d543a822488effca7","observation_id":"5f338e73-0c3a-4091-896c-6042d818d880","resolution":{"observed_at":"2026-08-01T22:35:51.781237Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1808.08929","last_updated":"2018-08-27T17:05:50Z","snapshot_observed_at":"2026-08-08T12:30:32.114433Z","submitted_at":"2018-08-27T17:05:50Z","title":"A neural attention model for speech command recognition","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1808.08929","snapshot_observed_at":"2026-08-01T22:35:51.929791Z","title":"A neural attention model for speech command recognition.arXiv preprint arXiv:1808.08929, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:51.929791Z"},"links":{"cited_paper":"/paper/1808.08929","citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:b61d7d903c705747d124059d2791deeba6a113e3db6f115516a2fa53a520f02e","observation_id":"c9ebb324-8bed-4ab6-b9fc-65827dcc2e29","resolution":{"observed_at":"2026-08-01T22:35:51.929791Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T22:35:52.025790Z","title":"An embarrassingly simple approach for trojan attack in deep neural networks","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-01T22:35:52.025790Z"},"links":{"citing_paper":"/paper/2607.15697"},"observation_digest":"sha256:7013d32935d1d624591ffa2198c01026a2aec2d054da0b38b571543e00717c70","observation_id":"c811794c-22f8-4c25-8785-5fd6659c016b","resolution":{"observed_at":"2026-08-01T22:35:52.025790Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2607.15697","last_updated":"2026-07-17T07:16:57Z","latest_version":1,"primary_category":"cs.SD","snapshot_observed_at":"2026-08-08T12:30:55.731497Z","submitted_at":"2026-07-17T07:16:57Z","title":"SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models"},"reference_resolution":{"displayed":44,"state_counts":{"malformed_identifier":2,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":42,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":44},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-08T06:32:00.761636+00:00","source":"crossref"},{"observed_at":"2026-08-08T06:31:55.24221+00:00","source":"retraction_watch"}],"thesis":"As of 9 August 2026, this Paper Citation Record lists 44 of 44 outbound references and 0 inbound Pith citation observations for arXiv:2607.15697."}