I’m running into some problems incorporating the DeepSpeech tensorflow lite model into my Android application. In particular, I’m getting a null pointer exception when I call tflite.run(input,output)
I was wondering if I could get some clarification on what I need to feed into the interpreter (i.e. what exactly the input and output should be and their dimensions)? Is there an example anywhere of running the tensorflow lite model in Android? I’m currently following the tutorial here (code) and have an audio recording in a float buffer array, but I’m not sure how to feed it into the model.
DeepSpeech version: 0.5.0
Tensorflow Lite version: 0.0.0-nightly