Phonexia, a spin-off company of the Brno University of Technology, has been pushing the limits of voice and speech recognition technologies for more than 13 years. Its technologies leverage the latest advancements in acoustics, phonetics, voice biometrics, and artificial intelligence to provide law enforcement and intelligence agencies with cutting-edge solutions to speaker identification and topic detection in large amounts of audio.

Phonexia stand (D 026) at Milipol Paris 2019 (Picture source: Army Recognition)

Phonexia Deep EmbeddingsSpeaker Identification technology is the world’s first commercially available voice biometrics engine powered entirely by deep neural networks making language-independent speaker identification incredibly fast and accurate. A recent independent study done by forensic experts at Bundeskriminalamt has acknowledged it to be the most accurate voice biometrics engine currently available on the market.

Whether you need to identify speakers in large amounts of audio, recognize a language and dialect of a speaker, detect topics, keywords, and another context in large quantities of audio recordings, and compare a speaker’s voice in forensic labs, Phonexia Speech Platform and its state-of-the-art speech and voice recognition technologies can solve it.