Modify Deepspeech-client after modifying audiofile_to_features

Hi,
As part of experiments we modified audiofile_to_features function and extracted features using a different method other than Mfcc.The training is completed and the model .pb file is also generated. For real-time inference deepSpeech Library calls SpeechToText function which is also to be modified in our case. I am not able to find how to modify it.
I will be thankful for any help in this.

Thanks and Regards,
Shilpa

native_client/deepspeech.cc ?