Thanks for tuning in to Google I/O. View all sessions on demandWatch on demand

Module: tflite_model_maker.audio_classifier

APIs to train an audio classification model.

Tutorial:

https://colab.research.google.com/github/googlecodelabs/odml-pathways/blob/main/audio_classification/colab/model_maker_audio_colab.ipynb

Demo code:

https://github.com/tensorflow/examples/blob/master/tensorflow_examples/lite/model_maker/demo/audio_classification_demo.py

Classes

class AudioClassifier: Audio classifier for training/inference and exporing.

class BrowserFftSpec: Model good at detecting speech commands, using Browser FFT spectrum.

class DataLoader: DataLoader for audio tasks.

class YamNetSpec: Model good at detecting environmental sounds, using YAMNet embedding.

Functions

create(...): Loads data and retrains the model.