TensorFlowASR ⚡

Almost State-of-the-art Automatic Speech Recognition in Tensorflow 2

TensorFlowASR implements some automatic speech recognition architectures such as DeepSpeech2, Conformer, etc. These models can be converted to TFLite to reduce memory and computation for deployment 😄

What's New?

• (11/14/2020) Supported Gradient Accumulation for Training in Larger Batch Size
• (11/3/2020) Reduce differences between librosa.stft and tf.signal.stft
• (10/31/2020) Update DeepSpeech2 and Supported Jasper https://arxiv.org/abs/1904.03288
• (10/18/2020) Supported Streaming Transducer https://arxiv.org/abs/1811.06621
• (10/15/2020) Add gradients accumulation and Refactor to TensorflowASR
• (10/10/2020) Update documents and upload package to pypi
• (10/6/2020) Change nlpaug version to >=1.0.1
• (9/18/2020) Support word-pieces (aka subwords) using tensorflow-datasets
• Support transducer tflite greedy decoding (conversion and invocation)
• Distributed training using tf.distribute.MirroredStrategy

Table of Contents

• What's New?
• Table of Contents
• 😋 Supported Models
• Installation
  • Installing via PyPi
  • Installing from source
• Setup training and testing
• TFLite Convertion
• Features Extraction
• Augmentations
• Training & Testing
• Corpus Sources and Pretrained Models
  • English
  • Vietnamese
  • German
• References & Credits

😋 Supported Models

• CTCModel (End2end models using CTC Loss for training)
• Deep Speech 2 (Reference: https://arxiv.org/abs/1512.02595) See examples/deepspeech2
• Jasper (Reference: https://arxiv.org/abs/1904.03288) See examples/jasper
• Transducer Models (End2end models using RNNT Loss for training)
• Conformer Transducer (Reference: https://arxiv.org/abs/2005.08100) See examples/conformer
• Streaming Transducer (Reference: https://arxiv.org/abs/1811.06621) See examples/streaming_transducer

Installation

Install tensorflow>=2.3.0 or tf-nightly.

For training and testing, you should use git clone for installing necessary packages from other authors (ctc_decoders, rnnt_loss, etc.)

Installing via PyPi

Run pip3 install -U TensorFlowASR

Installing from source

git clone https://github.com/TensorSpeech/TensorFlowASR.git
cd TensorFlowASR
python3 setup.py install

For anaconda3:

conda create -y -n tfasr tensorflow-gpu python=3.7 # tensorflow if using CPU
conda activate tfasr
pip install -U tensorflow-gpu # upgrade to latest version of tensorflow 
git clone https://github.com/TensorSpeech/TensorFlowASR.git
cd TensorFlowASR
python setup.py install

Setup training and testing

• For datasets, see datasets

• For training, testing and using CTC Models, run ./scripts/install_ctc_decoders.sh

• For training Transducer Models, run export CUDA_HOME=/usr/local/cuda && ./scripts/install_rnnt_loss.sh (Note: only export CUDA_HOME when you have CUDA)

• For mixed precision training, use flag --mxp when running python scripts from examples

• For enabling XLA, run TF_XLA_FLAGS=--tf_xla_auto_jit=2 python3 $path_to_py_script)

TFLite Convertion

After converting to tflite, the tflite model is like a function that transforms directly from an audio signal to unicode code points, then we can convert unicode points to string.

1. Install tf-nightly using pip install tf-nightly
2. Build a model with the same architecture as the trained model (if model has tflite argument, you must set it to True), then load the weights from trained model to the built model
3. Load TFSpeechFeaturizer and TextFeaturizer to model using function add_featurizers
4. Convert model's function to tflite as follows:

func = model.make_tflite_function(greedy=True) # or False
concrete_func = func.get_concrete_function()
converter = tf.lite.TFLiteConverter.from_concrete_functions([concrete_func])
converter.experimental_new_converter = True
converter.optimizations = [tf.lite.Optimize.DEFAULT]
converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS,
                                       tf.lite.OpsSet.SELECT_TF_OPS]
tflite_model = converter.convert()

5. Save the converted tflite model as follows:

if not os.path.exists(os.path.dirname(tflite_path)):
    os.makedirs(os.path.dirname(tflite_path))
with open(tflite_path, "wb") as tflite_out:
    tflite_out.write(tflite_model)

5. Then the .tflite model is ready to be deployed

Features Extraction

See features_extraction

Augmentations

See augmentations

Training & Testing

Example YAML Config Structure

speech_config: ...
model_config: ...
decoder_config: ...
learning_config:
  augmentations: ...
  dataset_config:
    train_paths: ...
    eval_paths: ...
    test_paths: ...
    tfrecords_dir: ...
  optimizer_config: ...
  running_config:
    batch_size: 8
    num_epochs: 20
    outdir: ...
    log_interval_steps: 500

See examples for some predefined ASR models and results

Corpus Sources and Pretrained Models

For pretrained models, go to drive

English

Name	Source	Hours
LibriSpeech	LibriSpeech	970h
Common Voice	https://commonvoice.mozilla.org	1932h

Vietnamese

Name	Source	Hours
Vivos	https://ailab.hcmus.edu.vn/vivos	15h
InfoRe Technology 1	InfoRe1 (passwd: BroughtToYouByInfoRe)	25h
InfoRe Technology 2 (used in VLSP2019)	InfoRe2 (passwd: BroughtToYouByInfoRe)	415h

German

Name	Source	Hours
Common Voice	https://commonvoice.mozilla.org/	750h

References & Credits

1. NVIDIA OpenSeq2Seq Toolkit
2. https://github.com/noahchalifour/warp-transducer
3. Sequence Transduction with Recurrent Neural Network
4. End-to-End Speech Processing Toolkit in PyTorch