top of page


High-performance, easy to power far-field voice embedded products
Excellent professional ability to provide high quality services.
Provide a total solution from acoustic design to voice recognition service docking.
Power manufacturers to produce and test voice products.

DSFF-020 Off-line far-field

voice recognition module

The DSFF-020 offline far-field speech recognition module has a

built-in NPU, supports quick customization of wake-up words, supports customization of offline control commands, and supports  voice recognition in Chinese, English, Korean, and Japanese. Rich

interfaces enable quickly embedded in products to make product

voice intelligence


Product diagram




Smart Home


Home appliances

Product advantage

Hardware advantages

High performance MCU

  • CPU main frequency 160MH, built-in SRAM up to 512KB.

  • BNPU upgrade, algorithm performance is greatly improved.

High integration and low solution cost

  • Built-in stereo audio codec (Audio Codec).

  • Rich interfaces: UART/I²S/SDIO/PWM/SPI/ I²C.

  • New generation process technology, lower power consumption.

Software advantages

  • Support voice recognition.

  • Support 2 microphone array.

  • Support AEC echo cancellation.

  • Support adaptive filtering.

  • Support offline command words with large vocabulary.

  • Support audio decoding, voice broadcast.

Product document

product document.png

Product brief

product document.png

User Manual

bottom of page