Microsoft Research's Speech Technology Group recently posted a short video demostrating how speech detection and recognition can be improved using a prototype hardware device that combines combines both a regular microphone and a bone-conductive microphone. Published research documents provide further details about the device and the benefits gained through its use along with improved processing algorithms:

"The device looks like a regular headset and it can be plugged into any machine with a USB port. The bone-conductive microphone has an interesting property: it is insensitive to ambient noise and captures the low frequency portion of the speech signals. Thanks to the signals from the boneconductive microphone, we are able to detect very robustly whether the speaker is talking, eliminating more than 90% of background speech. Furthermore, by combining both channels, we are able to significantly remove background speech even when the background speaker speaks at the same time as the speaker wearing the headset."

Some possible impacts of the new hardware and software combination include improved recognition accuracy in noisy environments, elimination of "Push-to-talk" interaction, and reduced bandwidth requirements for audio conferencing.

The demonstration video (along with other videos from the Speech Technology Group) is available from the group's Speech Videos page (last link on the page).

The research publications

• Multi-sensory Microphones for Robust Speech Detection, Enhancement, and Recognition
• Air-And Bone-Conductive Integrated Microphones for Robust Speech Detection and Enhancement