Hi DSP Concept Technical Support,
We are developing a special kind of multi-function smart speaker that utilizes the AVS and GVA on the same platform. In our use case which is a very small low-cost device (smaller than Amazon Echo Flex), we could not go for such a large Linux platform like quad-core Cortex A53/A35 or i.MX8, instead we have to stick with a cost-effective smaller footprint like STM32MP1 (dual-core Cortex A7+M4) Linux platform.
As per your advertisement on the TalkTo solution (https://w.dspconcepts.com/talkto) that is capable of supporting both AVS and GVA. So we would like to ask you if it is possible to implement both AVS and GVA on a single STM32MP1 platform (one at a time) or not? Is there any limitation in hardware?
We have already contacted to Siana-Systems (www.siana-systems.com) on this topic but apart from the AVS, they have no solution for the GVA. As stated by them the cloud connection responsible by the dual-core A7 @650-800MHz is not an issue, but this is due to some limitations in the M4 @209MHz of the STM32MP1 which is used to perform the far-field audio capture is not meet the GVA audio requirements. So is it possible to move the voice capture section to the dual-core A7 instead so as to gain enough processing power for both AVS and GVA requirements?
Or if you have any other preferences. Please kindly let us know in return.