IDEA #7GPW9L Websten System For Spontaneous Speech To Text Conversion

Areas of Application Web system for speech-to-text conversion of audio and video fi les loaded from user PC or from the Internet Specifi cation Languages: Ukrainian, English, and Russian (expansion is possible); word accuracy is 60-90% (depending on the record quality); information width of the primary audio stream: at least, 60 kbit/s Advantages Foreign analogs exist for other languages; speaker diarization; search of fragments by text, speaker, etc.; control of recognition speed depending on priority; text view synchronized with media content and vice versa; editing of the recognition results to obtain verifi ed transcripts Stage of Development. Suggestions for Commercialization IRL3, TRL5 Potential users: TV and radio studios, think tanks, intelligence services. Customizable upon request IPR Protection IPR1, IPR2