Daft-Exprt - PyTorch Implementation

PyTorch Implementation of Daft-Exprt: Robust Prosody Transfer Across Speakers for Expressive Speech Synthesis

The validation logs up to 70K of synthesized mel and alignment are shown below (VCTK_val_p237-088).


DATASET refers to the names of datasets such as VCTK in the following documents.


You can install the Python dependencies with

pip3 install -r requirements.txt

Also, Dockerfile is provided for Docker users.


You have to download the pretrained models and put them in output/ckpt/DATASET/.

For a multi-speaker TTS, run

python3 synthesize.py --text "YOUR_DESIRED_TEXT" --speaker_id SPEAKER_ID --restore_step RESTORE_STEP --mode single --dataset DATASET --ref_audio REF_AUDIO

to synthesize speech with the style of input audio at REF_AUDIO. The dictionary of learned speakers can be found at preprocessed_data/VCTK/speakers.json, and the generated utterances will be put in output/result/.

Batch Inference

Batch inference is also supported, try

python3 synthesize.py --source preprocessed_data/DATASET/val.txt --restore_step RESTORE_STEP --mode batch --dataset DATASET

to synthesize all utterances consuming themselves as a reference audio in preprocessed_data/DATASET/val.txt.


The pitch/volume/speaking rate of the synthesized utterances can be controlled by specifying the desired pitch/energy/duration ratios.
For example, one can increase the speaking rate by 20 % and decrease the volume by 20 % by

python3 synthesize.py --text "YOUR_DESIRED_TEXT" --speaker_id SPEAKER_ID --restore_step RESTORE_STEP --mode single --dataset DATASET --ref_audio REF_AUDIO --duration_control 0.8 --energy_control 0.8



The supported datasets are

  • VCTK: The CSTR VCTK Corpus includes speech data uttered by 110 English speakers (multi-speaker TTS) with various accents. Each speaker reads out about 400 sentences, which were selected from a newspaper, the rainbow passage and an elicitation paragraph used for the speech accent archive.
  • Any of multi-speaker TTS dataset (e.g., LibriTTS) can be added following VCTK.


  • For a multi-speaker TTS with external speaker embedder, download ResCNN Softmax+Triplet pretrained model of philipperemy's DeepSpeaker for the speaker embedding and locate it in ./deepspeaker/pretrained_models/.

  • Run

    python3 prepare_align.py --dataset DATASET

    for some preparations.

    For the forced alignment, Montreal Forced Aligner (MFA) is used to obtain the alignments between the utterances and the phoneme sequences.
    Pre-extracted alignments for the datasets are provided here.
    You have to unzip the files in preprocessed_data/DATASET/TextGrid/. Alternately, you can run the aligner by yourself.

    After that, run the preprocessing script by

    python3 preprocess.py --dataset DATASET


Train your model with

python3 train.py --dataset DATASET



tensorboard --logdir output/log

to serve TensorBoard on your localhost.
The loss curves, synthesized mel-spectrograms, and audios are shown.

Implementation Issues

  • RangeParameterPredictor is built with BiLSTM rather than a single linear layer with softplus() activation (it is however implemented and named as 'range_param_predictor_paper' in GaussianUpsampling).
  • Use 32 batch size instead of 48 due to memory issues.
  • Use log duration instead of normal duration.
  • Follow FastSpeech2 for the preprocess of pitch and energy.
  • Two options for embedding for the multi-speaker TTS setting: training speaker embedder from scratch or using a pre-trained philipperemy's DeepSpeaker model (as STYLER did). You can toggle it by setting the config (between 'none' and 'DeepSpeaker').
  • DeepSpeaker on VCTK dataset shows clear identification among speakers. The following figure shows the T-SNE plot of extracted speaker embedding.

  • For vocoder, HiFi-GAN and MelGAN are supported.


  author = {Lee, Keon},
  title = {Daft-Exprt},
  year = {2021},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/keonlee9420/Daft-Exprt}}