Deep Speech features

Hello. Tell me where you can watch the video or a description to read what happens to the signal in the steps before calculating the signs. How are the signs and features of the recognized signal calculated on what principles. For example, as in my system ERS VCRS

What algorithms and model are used after identifying features and training the neural network. All this I need to understand what is the difference from other speech recognition systems such as CMU Sphinx. For a quick start and to improve the volume of experiments.

Sorry, but have you just had a look at the documentation? The model is explained, and I think all your questions are already answered in that.

Like that https://deepspeech.readthedocs.io/en/v0.6.0/DeepSpeech.html ?