A ReSpeaker Core v2.0 Walks Into a Bar…

… and we welcome a new family member to the team!
ReSpeaker Core v2.0
Hey Seeekers,

We are back with another addition to our ReSpeaker product line – the ReSpeaker Core v2.0. The ReSpeaker Core v2.0 is geared towards helping create voice interaction projects. Featuring a powerful computing core and a 6-mic array with professional voice enhancement algorithms, this board is a significant upgrade from the original ReSpeaker Core we launched on Kickstarter back in 2016. The board runs Debian OS, and as of today is compatible with Android 8.1, allowing developers flexibility in their projects. Oh – and check below for some videos!

ReSpeaker Core v2.0
The ReSpeaker Core v2.0 is powerful enough to support 4k video output via HDMI, which enables it to act as a rich multimedia platform. This is possible thanks to the RK3229, a quad-core processor with a dedicated GPU, and 1GB of RAM. The OS can be booted from the onboard 4GB eMMC, or through the microSD slot.

The previous ReSpeaker products either had no voice enhancement algorithms or relied on the XMOS chipsets to provide these. This is not the case with the ReSpeaker Core v2.0 as the voice enhancement algorithms are now software based. In this way, developers can create their own, or chose the algorithms best suited for them. By default, the board uses some algorithms provided by Alango, a professional audio DPS company trusted and used in the automotive industry. These algorithms can be accessed through a wrapper library to allow developers the choice of when to use them.

We recognize that every project has different requirements, so with the ReSpeaker Core v2.0 we are enabling developers with as many options as possible, from the algorithms, NLP services, smart assistants, and operating systems. If you have suggestions on what services you would like to see us officially support, let us know on our forums, or if you want to port it over yourself, submit a merge request to our GitHub.

The ReSpeaker Core v2.0 is also able to be customized. The design of the device enables quick customization of peripherals such as adding or removing connectors, increasing eMMC capacity, or reconfiguring the mic array. If you have questions regarding customization services please send them to [email protected].

Smart Receptionist Demo


Alexa Demo

Wake Up Distance Test

Note: In this test the distance achieved was quite extreme due to the environment.

We ran a test in our other office. Here are the results:

Voice Volume ~85dB
Background Volume ~60dB
Distance (m) Pass/Tests Rate(%)
1 10/10 100
2 9/10 90
3 9/10 90
4 9/10 90
5 8/10 80
6 10/10 100
8 9/10 90
10 5/10 50
12 1/10 10

About Author


June 2018