pre-trained graph for Voice recognition?

RTasa · ‎12-13-2018

Is there a pre-trained graph for Open Vino / Movidius for Voice recognition ?

nikos1 · ‎12-13-2018

Hi Bob,

the only pre-trained ASR I could find was in https://download.01.org/openvinotoolkit/2018_R3/models_contrib/GNA/

but probably not good for Movidius (?). I believe speech on CPU, GPU or GNA only.

Cheers,

Nikos

RTasa · ‎12-14-2018

Thanks, I hope Intel can create this model that is used with the Movidius chip set. That would be HUGE. We have all kinds of models for seeing people detection but interacting with them is very limited,. Actual speech and understanding of the spoken language is very limited on the edge lead to true AI experience.

nikos1 · ‎12-14-2018

Very good point. Facing similar issue here. GNA will help a bit in this direction.