i.MX 8M Mini Development Kit for Amazon Alexa Voice Service
IMX8MMINIALEXADEVKIT Receive alerts
NXP has partnered with Sugr to create the i.MX 8M Mini Development kit for Amazon Alexa Voice Service (AVS) to enable developers to easily integrate AVS into their next generation design. Powered by the i.MX 8M Mini applications processor, the Sugr Sense N software stack includes a full far field audio front end (omni-directional wake up, sound localization, beamforming, noise reduction, reverberation removal, acoustic echo cancellation, full-duplex voice communication) and Alexa Voice Service features such as Alexa Wake Word, Multi-Room-Music and Alexa Call Manager.
The kit includes the NXP 8MIC-RPI-MX8 board that supports several microphone configurations including 2-Mic linear, 3-Mic linear and circular, 4-Mic Trillium and rectangular, and 6-Mic circular. This flexibility along with the full featured Sugr Sense N software enables developers to create high-performance far-field designs for a wide range of target applications.
To request a kit please send an e-mail to: NXP-AVSDevKit@nxp.com.
Product Details
Supported Devices
Processors and Microcontrollers
i.MX 8 Applications Processors
- i.MX8MMINI: i.MX 8M Mini - Arm® Cortex®-A53, Cortex-M4, Audio, Voice, Video
Features
Processor
- i.MX 8M Mini Applications Processor
- 4x Cortex-A53 @ 1.8 GHz
- 1x Cortex-M4 @ 400 MHz
- KM35Z512VLQ7 MC
Microphone Configurations
- 2-Mic linear
- 4-Mic linear
- 4-Mic circular arrays
Software Support
- Sugr Cloud services for device management and firmware OTA
- Quick integration with Alexa Voice Service and provide pre-test for Amazon AVS far-field certification
Connectivity
- Wi-Fi 802.11 a/b/g/n/ac max. data rate 433Mbps
- Bluetooth 4.1+2.1EDR, support A2DP, AVRCP, HFP Profile
- Smart BLE and Soft AP network configuration
Other Features
- Full-duplex voice communication with telecom level low latency
- Mic-array customization - flexible for ID and MD design
- multi-channel AEC, noice reduction,beamforming, reverberation removal, etc.
- 360° sound source localization (precision ± 10°)
- Far-field wake up and voice pick-up
- Wake up precisely 5M FRR (Wake Word False Rejection Rate) < 10% (65dBc noise environment) , 24 hours FAR (Wake Word False Alarm Rate) < 3 times
- High Response Accuracy Rate (RAR), 5M RAR> 85%, 3M RAR>90% (device playback mode)