This repository contains code for synthesizing speech audio from silently mouthed words captured with electromyography (EMG). It is the official repository for the papers Digital Voicing of Silent Speech at EMNLP 2020, An Improved Model for Voicing Silent Speech at ACL 2021, and the dissertation Voicing Silent Speech. The current commit contains only the most recent model, but the versions from prior papers can be found in the commit history. On an ASR-based open vocabulary evaluation, the latest model achieves a WER of approximately 36%. Audio samples can be found here.
The repository also includes code for directly converting silent speech to text. See the section labeled Silent Speech Recognition.
The EMG and audio data can be downloaded from https://doi.org/10.5281/zenodo.4064408. The scripts expect the data to be located in a emg_data subdirectory by default, but the location can be overridden with flags (see the top of read_emg.py).
Force-aligned phonemes from the Montreal Forced Aligner have been included as a git submodule, which must be updated using the process described in "Environment Setup" below. Note that there will not be an exception if the directory is not found, but logged phoneme prediction accuracies reporting 100% is a sign that the directory has not been loaded correctly.
We strongly recommend running in Anaconda. To create a new environment with all required dependencies, run:
conda env create -f environment.yml
conda activate silent_speech
This will install with CUDA 11.8.
You will also need to pull git submodules for Hifi-GAN and the phoneme alignment data, using the following commands:
git submodule init
git submodule update
tar -xvzf text_alignments/text_alignments.tar.gz
Use the following commands to download pre-trained DeepSpeech model files for evaluation. It is important that you use DeepSpeech version 0.7.0 model files for evaluation numbers to be consistent with the original papers. Note that more recent DeepSpeech packages such as version 0.9.3 can be used as long as they are compatible with version 0.7.x model files.
curl -LO https://github.com/mozilla/DeepSpeech/releases/download/v0.7.0/deepspeech-0.7.0-models.pbmm
curl -LO https://github.com/mozilla/DeepSpeech/releases/download/v0.7.0/deepspeech-0.7.0-models.scorer
(Optional) Training will be faster if you re-run the audio cleaning, which will save re-sampled audio so it doesn't have to be re-sampled every training run.
python data_collection/clean_audio.py emg_data/nonparallel_data emg_data/silent_parallel_data emg_data/voiced_parallel_data
Pre-trained models for the vocoder and transduction model are available at https://doi.org/10.5281/zenodo.6747411.
To train an EMG to speech feature transduction model, use
python transduction_model.py --hifigan_checkpoint hifigan_finetuned/checkpoint --output_directory "./models/transduction_model/"
where hifigan_finetuned/checkpoint is a trained HiFi-GAN generator model (optional).
At the end of training, an ASR evaluation will be run on the validation set if a HiFi-GAN model is provided.
To evaluate a model on the test set, use
python evaluate.py --models ./models/transduction_model/model.pt --hifigan_checkpoint hifigan_finetuned/checkpoint --output_directory evaluation_output
By default, the scripts now use a larger validation set than was used in the original EMNLP 2020 paper, since the small size of the original set gave WER evaluations a high variance. If you want to use the original validation set you can add the flag --testset_file testset_origdev.json.
The HiFi-GAN model is fine-tuned from a multi-speaker model to the voice of this dataset. Spectrograms predicted from the transduction model are used as input for fine-tuning instead of gold spectrograms. To generate the files needed for HiFi-GAN fine-tuning, run the following with a trained model checkpoint:
python make_vocoder_trainset.py --model ./models/transduction_model/model.pt --output_directory hifigan_training_files
The resulting files can be used for fine-tuning using the instructions in the hifi-gan repository.
The pre-trained model was fine-tuned for 75,000 steps, starting from the UNIVERSAL_V1 model provided by the HiFi-GAN repository.
Although the HiFi-GAN is technically fine-tuned for the output of a specific transduction model, we found it to transfer quite well and shared a single HiFi-GAN for most experiments.
This section is about converting silent speech directly to text rather than synthesizing speech audio. The speech-to-text model uses the same neural architecture but with a CTC decoder, and achieves a WER of approximately 28% (as described in the dissertation Voicing Silent Speech).
You will need to install the ctcdecode library (1.0.3) in addition to the libraries listed above to use the recognition code. (This package cannot be built successfully under Windows platform)
pip install git+https://github.com/parlance/ctcdecode.git
And you will need to download a KenLM language model, such as this one from DeepSpeech:
curl https://github.com/mozilla/DeepSpeech/releases/download/v0.6.1/lm.binary
Pre-trained model weights can be downloaded from https://doi.org/10.5281/zenodo.7183877.
To train a model, run
python recognition_model.py --output_directory "./models/recognition_model/"
To run a test set evaluation on a saved model, use
python recognition_model.py --evaluate_saved "./models/recognition_model/model.pt"