Streaming speech recognition on the STM32N6 NPU: could you evaluate my result?

2 replies
97 views

I have implemented streaming speech recognition on the STM32N6, with a model designed specifically for the constraints of this microcontroller and its NPU.

The STM32N6570-DK listens to its on-board microphone and prints the text on the display. It recognizes any English words, not a fixed list of commands, and works with no network connection. The word error rate on LibriSpeech test-other is 14.84% with no language model.

Source code, model weights and ready-to-flash firmware: https://github.com/fotonik-sk/stm32n6-asr

Maxime MARCHETTOBest answer

ST Community Manager

Hello @fotonik-sk,

Thank you for sharing your project!

To help others interested in your work find this post more easily, I’ve marked my reply as the solution.

Regards,
Maxime

I have implemented streaming speech recognition on the STM32N6, with a model designed specifically for the constraints of this microcontroller and its NPU.
The STM32N6570-DK listens to its on-board microphone and prints the text on the display. It recognizes any English words, not a fixed list of commands, and works with no network connection. The word error rate on LibriSpeech test-other is 14.84% with no language model.

Source code, model weights and ready-to-flash firmware: https://github.com/fotonik-sk/stm32n6-asr

In order to give better visibility on the answered topics, please click on 'Best answer' on the reply which solved your issue or answered your question.

L

LMSTM

ST Employee

Hello, very cool demo ! You could probably see a significant speedup if you manage to shrink your model so that all activations fit in internal SRAM.

I think a very interesting part of your work is also the 248 hours of speech recorded using the on-board microphone. I do not know if you are at liberty to release it publicly, but it would be very useful for the community !