How to do Contact Name Recognition using DeepSpeech

DeepSpeech
#1

I was wondering if there is a way to handle Contact Names using DeepSpeech ?

For example

Call Bob or Please ask Rahul to call me up

Based on my search I found the following paper from Google

[Personalized Speech Recognition on Mobile Devices](http://www.mirlab.org/conference_papers/international_conference/ICASSP%202016/pdfs/0005955.pdf)

Which says to make use of Class based Language models along with Finite State Transducers.

Any idea how could we get this done using DeepSpeech or if at all I’m on the right path

#2

If you’re looking to implement this with a pre-trained model, using a class-based language model with the new decoder works reasonably well. The pre-trained model doesn’t include a lot of proper names in it though, and it’s too slow for real-time inference on mobile devices as it’s currently implemented. If you’re training from your own audio, a smaller model (1024 or 1536 wide) can be real-time on a Snapdragon 835 for example, and paired with a class-based LM it should work reasonably well.

#3

Hey Reuben,

Thanks for your response. I do not intend to develop it for mobile devices. I am currently training from scratch on the Common Voice corpus.

My concern was whether it is achievable (as per your response it is a yes). Could you please point to resources to which I can refer to get it done?

#4

Can you please expand on class based LM. Is it possible to do using kenlm tool?