Howl

Wake word detection modeling for Firefox Voice, supporting open datasets like Google Speech Commands and Mozilla Common Voice.

Citation:

@inproceedings{tang-etal-2020-howl,
    title = "Howl: A Deployed, Open-Source Wake Word Detection System",
    author = "Tang, Raphael and Lee, Jaejun and Razi, Afsaneh and Cambre, Julia and Bicking, Ian and Kaye, Jofish and Lin, Jimmy",
    booktitle = "Proceedings of Second Workshop for NLP Open Source Software (NLP-OSS)",
    month = nov,
    year = "2020",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/2020.nlposs-1.9",
    doi = "10.18653/v1/2020.nlposs-1.9",
    pages = "61--65"
}

Quickstart Guide

Install PyAudio and PyTorch 1.5+ through your distribution's package system.
Install Howl using pip

pip install howl

To immediately use a pre-trained Howl model for inference, we provide the client API. The following example (also found under examples/hey_fire_fox.py) loads the "hey_fire_fox" pretrained model with a simple callback and starts the inference client.

from howl.client import HowlClient

def hello_callback(detected_words):
    print("Detected: {}".format(detected_words))

client = HowlClient()
client.from_pretrained("hey_fire_fox", force_reload=False)
client.add_listener(hello_callback)
client.start().join()

Training Guide

Installation

git clone https://github.com/castorini/howl && cd howl
Install PyTorch by following your platform-specific instructions.
Install PyAudio and its dependencies through your distribution's package system.
pip install -r requirements.txt -r requirements_training.txt (some apt packages might need to be installed)

Preparing a Dataset

In the example that follows, we describe how to train a custom detector for the word, "fire."

Download a supported data source. We recommend Common Voice for its breadth and free license.
To provide alignment for the data, install Montreal Forced Aligner (MFA) and download an English pronunciation dictionary.
Create a positive dataset containing the keyword:

VOCAB='["fire"]' INFERENCE_SEQUENCE=[0] DATASET_PATH=data/fire-positive python -m training.run.create_raw_dataset -i ~/path/to/common-voice --positive-pct 100 --negative-pct 0

Create a negative dataset without the keyword: note that 5% is sufficient when generating negative dataset from common-voice dataset

VOCAB='["fire"]' INFERENCE_SEQUENCE=[0] DATASET_PATH=data/fire-negative python -m training.run.create_raw_dataset -i ~/path/to/common-voice --positive-pct 0 --negative-pct 5

Generate some mock alignment for the negative set, where we don't care about alignment:

DATASET_PATH=data/fire-negative python -m training.run.attach_alignment --align-type stub

Use MFA to generate alignment for the positive set:

mfa_align data/fire-positive/audio eng.dict pretrained_models/english.zip output-folder

Attach the MFA alignment to the positive dataset:

DATASET_PATH=data/fire-positive python -m training.run.attach_alignment --align-type mfa -i output-folder

Training and Running a Model

Source the relevant environment variables for training the res8 model: source envs/res8.env.
Train the model: python -m training.run.train -i data/fire-negative data/fire-positive --model res8 --workspace workspaces/fire-res8.
For the CLI demo, run python -m training.run.demo --model res8 --workspace workspaces/fire-res8.

Pretrained Models

howl-models contains workspaces with pretrained models

To get the latest models, simply run git submodule update --init --recursive

hey firefox

VOCAB='[" hey","fire","fox"]' INFERENCE_SEQUENCE=[0,1,2] INFERENCE_THRESHOLD=0 NUM_MELS=40 MAX_WINDOW_SIZE_SECONDS=0.5 python -m training.run.demo --model res8 --workspace howl-models/howl/hey-fire-fox

Reproducing Paper Results

First, follow the installation instructions in the quickstart guide.

Google Speech Commands

Download the Google Speech Commands dataset and extract it.
Source the appropriate environment variables: source envs/res8.env
Set the dataset path to the root folder of the Speech Commands dataset: export DATASET_PATH=/path/to/dataset
Train the res8 model: NUM_EPOCHS=20 MAX_WINDOW_SIZE_SECONDS=1 VOCAB='["yes","no","up","down","left","right","on","off","stop","go"]' BATCH_SIZE=64 LR_DECAY=0.8 LEARNING_RATE=0.01 python -m training.run.pretrain_gsc --model res8

Hey Firefox

Download the Hey Firefox corpus, licensed under CC0, and extract it.
Download our noise dataset, built from Microsoft SNSD and MUSAN, and extract it.
Source the appropriate environment variables: source envs/res8.env
Set the noise dataset path to the root folder: export NOISE_DATASET_PATH=/path/to/snsd
Set the firefox dataset path to the root folder: export DATASET_PATH=/path/to/hey_firefox
Train the model: LR_DECAY=0.98 VOCAB='[" hey","fire","fox"]' USE_NOISE_DATASET=True BATCH_SIZE=16 INFERENCE_THRESHOLD=0 NUM_EPOCHS=300 NUM_MELS=40 INFERENCE_SEQUENCE=[0,1,2] MAX_WINDOW_SIZE_SECONDS=0.5 python -m training.run.train --model res8 --workspace workspaces/hey-ff-res8

Hey Snips

Download hey snips dataset
Process the dataset to a format howl can load

VOCAB='["hey","snips"]' INFERENCE_SEQUENCE=[0,1] DATASET_PATH=data/hey-snips python -m training.run.create_raw_dataset --dataset-type 'hey-snips' -i ~/path/to/hey_snips_dataset

Generate some mock alignment for the dataset, where we don't care about alignment:

DATASET_PATH=data/hey-snips python -m training.run.attach_alignment --align-type stub

Use MFA to generate alignment for the dataset set:

mfa_align data/hey-snips/audio eng.dict pretrained_models/english.zip output-folder

Attach the MFA alignment to the dataset:

DATASET_PATH=data/hey-snips python -m training.run.attach_alignment --align-type mfa -i output-folder

Source the appropriate environment variables: source envs/res8.env
Set the noise dataset path to the root folder: export NOISE_DATASET_PATH=/path/to/snsd
Set the noise dataset path to the root folder: export DATASET_PATH=/path/to/hey-snips
Train the model: LR_DECAY=0.98 VOCAB='[" hey","snips"]' USE_NOISE_DATASET=True BATCH_SIZE=16 INFERENCE_THRESHOLD=0 NUM_EPOCHS=300 NUM_MELS=40 INFERENCE_SEQUENCE=[0,1] MAX_WINDOW_SIZE_SECONDS=0.5 python -m training.run.train --model res8 --workspace workspaces/hey-snips-res8

Generating dataset for Mycroft-precise

howl also provides a script for transforming howl dataset to mycroft-precise dataset

VOCAB='[" hey","fire","fox"]' INFERENCE_SEQUENCE=[0,1,2] python -m training.run.generate_precise_dataset --dataset-path /path/to/howl_dataset

Experiments

To verify the correctness of our implementation, we first train and evaluate our models on the Google Speech Commands dataset, for which there exists many known results. Next, we curate a wake word detection datasets and report our resulting model quality.

For both experiments, we generate reports in excel format. experiments folder includes sample outputs from the for each experiment and corresponding workspaces can be found here

commands_recognition

For command recognition, we train the four different models (res8, LSTM, LAS encoder, MobileNetv2) to detect twelve different keywords: “yes”, “no”, “up”, “down”, “left”, “right”, “on”, “off”, “stop”, “go”, unknown, or silence.

python -m training.run.eval_commands_recognition --num_iterations n --dataset_path < path_to_gsc_datasets >

word_detection

In this experiment, we train our best commands recognition model, res8, for hey firefox and hey snips and evaluate them with different threashold.

Two different performance reports are generated, one with the clean audio and one with audios with noise

python -m training.run.eval_wake_word_detection --num_models n --hop_size < number between 0 and 1 > --exp_type < hey_firefox | hey_snips > --dataset_path "x" --noiseset_path "y"

We also provide a script for generating ROC curve. exp_timestamp can be found from the reports generated from previous command

python -m training.run.generate_roc --exp_timestamp < experiment timestamp > --exp_type < hey_firefox | hey_snips >

wangwei2009/howl