Datasets
Buyer room
DOCS.PHONEXIA.COMManage supplier listing
Phonexia — Proprietary labelled speech-authenticity evaluation corpora, disclosed indirectly but decisively through the product's calibration statement: the Replay Attack Detection score is a Log-Likelihood Ratio whose zero point corresponds to the Equal Error Rate (EER) on their evaluation datasets, with practical scores falling between -4.5 and +1.0. Producing an EER-calibrated threshold requires a held-out corpus with per-clip genuine-versus-replay ground truth, so the vendor necessarily maintains a labelled authenticity evaluation set. Content spans the acoustic dimensions named as the de…
supplier index read from their site
Proprietary labelled speech-authenticity evaluation corpora, disclosed indirectly but decisively through the product's calibration statement: the Replay Attack Detection score is a Log-Likelihood Ratio whose zero point corresponds to the Equal Error Rate (EER) on their evaluation datasets, with practical scores falling between -4.5 and +1.0. Producing an EER-calibrated threshold requires a held-out corpus with per-clip genuine-versus-replay ground truth, so the vendor necessarily maintains a labelled authenticity evaluation set. Content spans the acoustic dimensions named as the detection features — room reverberation, microphone quality, background noise and transmission characteristics — implying paired live-versus-replayed recordings of the same speakers across differing capture and transmission conditions. Part of a wider Deepfake Detection suite, so synthesis labels likely accompany replay labels
From 12 yearsCoverage Information TechnologyAsset class Equities
Sample, licence terms, pricing and eval results when Phonexia publishes them. Until then, discover alternatives today with a 7-day trial.
Start 7-day trial →What buyers ask about Phonexia, answered from this page.
Phonexia — Proprietary labelled speech-authenticity evaluation corpora, disclosed indirectly but decisively through the product's calibration statement: the Replay Attack Detection score is a Log-Likelihood Ratio whose zero point corresponds to the Equal Error Rate (EER) on their evaluation datasets, with practical scores falling between -4.5 and +1.0. Producing an EER-calibrated threshold requires a held-out corpus with per-clip genuine-versus-replay ground truth, so the vendor necessarily maintains a labelled authenticity evaluation set. Content spans the acoustic dimensions named as the de…
Phonexia offers (Alternative, Reference) — The authenticity evaluation corpus is not sized publicly, but computing an Equal Error Rate implies a substantial held-out labelled set rather than a handful of samples — realistically tens of thousands of paired live-and-attack clips across the four named nuisance dimensions to support a stable EER estimate. Phonexia's wider speech portfolio (speaker recognition, language and speaker identification, transcription across multiple languages) implies a far larger aggregate speech corpus than the authenticity slice alone. Request further research for corpus sizes and language coverage.
Training and benchmarking audio deepfake, synthesis and replay detectors; voice-biometric liveness and anti-spoofing systems for banking apps and smart devices; call-centre fraud detection on replayed customer audio; building adversarial evaluation suites and standardised replay-attack benchmarks; channel-robustness and domain-shift testing for speaker-verification models; and threshold calibration for any voice-authentication deployment needing a defensible operating point
The data is with 12 years of history.
Coverage spans US, Other; Application Software; alternative, reference; equities.
Phonexia has not added their own details yet. Not yet on file: