Binbin Zhang

Thanks for your kind remind. The project is mainly based on https://github.com/jingyonghou/KWS_Max-pooling_RHE, which is published in "Mining Effective Negative Training Samples for Keyword Spotting". The author is jingyonghou, who is...

Is the model architecure exactly same to yours? I don't think so, we did a lot of experiments with different model size. Please refer "Hello Edge: Keyword Spotting on Microcontrollers"...

And all the parameter could be easily changed in yaml configure file. please see https://github.com/wenet-e2e/wenet-kws/blob/master/examples/hi_xiaowen/s0/conf/mdtc.yaml for demo.

Does it work for real hotword system?

The wakeup position is not accurate now, we are trying to figure it out.

Not yet, but it's in plan.

It's hard to say, but the most important things are: 1. your platform: you should choose proper inference runtime for the platform. 2. streamming: hotword is a streaming task in...

I have no bandwidth to do it recently.

There is no speaker information in the corpus.

please call decoder.reset() after calling decode_wav to reset decoding status.