Speech Recognition News: embedded vs. cloud, voice prosthesis, voice-to-email
On the one hand, some say speech recognition is dead. On the other, innovation in terms of where, when and how consumers and enterprises use speech recognition is constantly changing. The week in news and developments in the world of speech recognition:
- What is the future of speech recognition: embedded technologies or the cloud? David Burke, Chief of Engineering for Google, hedged in a SpeechTEK interview. He acknowledged, “Embedded technologies promise faster response times and are immune to patchy network conditions. On the other hand, cloud-based services offer significantly larger and more sophisticated language and acoustic models for speech recognition and higher quality text-to-speech.” So which is it? The answer, as always, is “it depends.”
- Voice prosthesis: just speak into the electrodes. NPR reported recently on a new speech technology that could replaced the robotic-sounding electrolarynx for those who have lost their voices due to throat cancer. The technology, a small valve surgically inserted into an opening in the throat, is manipulated by signals from electrodes attached to the face and by the user holding his thumb across the opening and forcing air upwards.
- May I take your email… while driving? BMW has announced a new speech recognition system that will allow drivers to dictate emails while driving.
- Medical speech recognition is here. Nuance announced its Nuance Healthcare Development Platform will offer users the possibility of “speech-driven clinical documentation, voice navigation and automatic clinical fact extraction.”