- January 2021: Release v1.2
- January 2021: Joint CTC-Attention Transformer model passing
- January 2021: Speech Transformer model passing
- January 2021: Apply Hydra: framework for elegantly configuring complex applications
- December 2020: Release v1.1
- December 2020: Update pre-train models
- December 2020: Joint CTC-Attention LAS passing (Currently, Not Supports Multi-GPU)
- November 2020: Deep Speech 2 passing
- November 2020: KsponSpeech Subword & Grapheme Unit passing (Not Tested)
- November 2020: RAdam & AdamP Optimizer passing
- Currently, beam search may not work properly.
- The model's inference is currently not working properly, but will be updated as soon as possible.
KoSpeech, an open-source software, is modular and extensible end-to-end Korean automatic speech recognition (ASR) toolkit based on the deep learning library PyTorch. Several automatic speech recognition open-source toolkits have been released, but all of them deal with non-Korean languages, such as English (e.g. ESPnet, Espresso). Although AI Hub opened 1,000 hours of Korean speech corpus known as KsponSpeech, there is no established preprocessing method and baseline model to compare model performances. Therefore, we propose preprocessing methods for KsponSpeech corpus and a baseline model for benchmarks. Our baseline model is based on Listen, Attend and Spell (LAS) architecture and ables to customize various training hyperparameters conveniently. By KoSpeech, we hope this could be a guideline for those who research Korean speech recognition. Our baseline model achieved 10.31% character error rate (CER) at KsponSpeech corpus only with the acoustic model.
Description | Feature | Dataset | Epochs | CER | Model |
---|---|---|---|---|---|
Deep Speech 2 | Kaldi-style fbank 80 | KsponSpeech | 1 | 42.3 | download |
Listen, Attend and Spell | Kaldi-style fbank 80 | KsponSpeech | 1 | 19.2 | download |
Joint CTC-Attention | Kaldi-style fbank 80 | KsponSpeech | 10 | 12.5 | download |
VAD Model | - | - | - | - | download |
※ Training is in progress. As the training progresses, the pre-trained model will be updated.
Dataset | Authentication | Output-Unit | Transcript |
---|---|---|---|
KsponSpeech | Required | Character | download |
KsponSpeech | Required | Subword | download |
KsponSpeech | Required | Grapheme | download |
LibriSpeech | Unrequired | Subword | download |
※ Authentication : Refer to [Link]
End-to-end (E2E) automatic speech recognition (ASR) is an emerging paradigm in the field of neural network-based speech recognition that offers multiple benefits. Traditional “hybrid” ASR systems, which are comprised of an acoustic model, language model, and pronunciation model, require separate training of these components, each of which can be complex.
For example, training of an acoustic model is a multi-stage process of model training and time alignment between the speech acoustic feature sequence and output label sequence. In contrast, E2E ASR is a single integrated approach with a much simpler training pipeline with models that operate at low audio frame rates. This reduces the training time, decoding time, and allows joint optimization with downstream processing such as natural language understanding.
-
Attention (Multi-Head / Location-Aware / Additive / Scaled-dot)
-
Various Feature (Spectrogram / Mel-Spectrogram / MFCC / Filter-Bank)
-
Scheduled Sampling (Teacher forcing scheduling)
-
Inference with batching
-
Multi-GPU training
We have referred to several papers to develop the best model possible. And tried to make the code as efficient and easy to use as possible. If you have any minor inconvenience, please let us know anytime.
We will response as soon as possible.
So far, Five models are implemented: Deep Speech 2, Listen Attend and Spell (LAS), Speech Transformer, Joint CTC-Attention LAS, and Joint CTC-Attention Transformer. To check details of these model architectures, check figures attached to each section.
- Deep Speech 2
Deep Speech 2 showed faster and more accurate performance on ASR tasks with Connectionist Temporal Classification (CTC) loss. This model has been highlighted for significantly increasing performance compared to the previous end- to-end models.
- Listen, Attend and Spell
We follow the architecture previously proposed in the "Listen, Attend and Spell", but some modifications were added to improve performance. We provide four different attention mechanisms, scaled dot-product attention
, additive attention
, location aware attention
, multi-head attention
. Attention mechanisms much affect the performance of models.
- Speech Transformer
Transformer is a powerful architecture in the Natural Language Processing (NLP) field. This architecture also showed good performance at ASR tasks. In addition, as the research of this model continues in the natural language processing field, this model has high potential for further development.
- Joint CTC-Attention
With the proposed architecture to take advantage of both the CTC-based model and the attention-based model. It is a structure that makes it robust by adding CTC to the encoder.
We support various options for training. More details please check here.
- Options
usage: main.py [-h] [--mode MODE] [--sample_rate SAMPLE_RATE]
[--frame_length FRAME_LENGTH] [--frame_shift FRAME_SHIFT]
[--n_mels N_MELS] [--normalize] [--del_silence]
[--input_reverse] [--feature_extract_by FEATURE_EXTRACT_BY]
[--transform_method TRANSFORM_METHOD]
[--freq_mask_para FREQ_MASK_PARA]
[--time_mask_num TIME_MASK_NUM] [--freq_mask_num FREQ_MASK_NUM]
[--architecture ARCHITECTURE] [--use_bidirectional]
[--mask_conv] [--hidden_dim HIDDEN_DIM] [--dropout DROPOUT]
[--num_heads NUM_HEADS] [--label_smoothing LABEL_SMOOTHING]
[--num_encoder_layers NUM_ENCODER_LAYERS]
[--num_decoder_layers NUM_DECODER_LAYERS] [--rnn_type RNN_TYPE]
[--extractor EXTRACTOR] [--activation ACTIVATION]
[--attn_mechanism ATTN_MECHANISM]
[--teacher_forcing_ratio TEACHER_FORCING_RATIO]
[--num_classes NUM_CLASSES] [--d_model D_MODEL]
[--ffnet_style FFNET_STYLE] [--dataset_path DATASET_PATH]
[--transcripts_path TRANSCRIPTS_PATH]
[--data_list_path DATA_LIST_PATH] [--spec_augment] [--use_cuda]
[--batch_size BATCH_SIZE] [--num_workers NUM_WORKERS]
[--num_epochs NUM_EPOCHS] [--init_lr INIT_LR]
[--peak_lr PEAK_LR] [--final_lr FINAL_LR]
[--final_lr_scale FINAL_LR_SCALE]
[--init_lr_scale INIT_LR_SCALE] [--max_len MAX_LEN]
[--max_grad_norm MAX_GRAD_NORM] [--weight_decay WEIGHT_DECAY]
[--reduction REDUCTION] [--warmup_steps WARMUP_STEPS]
[--teacher_forcing_step TEACHER_FORCING_STEP]
[--min_teacher_forcing_ratio MIN_TEACHER_FORCING_RATIO]
[--seed SEED] [--save_result_every SAVE_RESULT_EVERY]
[--checkpoint_every CHECKPOINT_EVERY]
[--print_every PRINT_EVERY] [--resume]
This project recommends Python 3.7 or higher.
We recommend creating a new virtual environment for this project (using virtual env or conda).
- Numpy:
pip install numpy
(Refer here for problem installing Numpy). - Pytorch: Refer to PyTorch website to install the version w.r.t. your environment.
- Pandas:
pip install pandas
(Refer here for problem installing Pandas) - Matplotlib:
pip install matplotlib
(Refer here for problem installing Matplotlib) - librosa:
conda install -c conda-forge librosa
(Refer here for problem installing librosa) - torchaudio:
pip install torchaudio==0.6.0
(Refer here for problem installing torchaudio) - tqdm:
pip install tqdm
(Refer here for problem installing tqdm) - sentencepiece:
pip install sentencepiece
(Refer here for problem installing sentencepiece) - hydra:
pip install hydra-core --upgrade
(Refer here for problem installing hydra)
Currently we only support installation from source code using setuptools. Checkout the source code and run the
following commands:
pip install -e .
Download from here or refer to the following to preprocess.
- KsponSpeech : Check this page
- LibriSpeech : Check this page
You can choose from four models and training this. There are many other training options, so look carefully and execute the following command:
- Deep Speech 2 Training
python ./bin/main.py model=ds2 train=ds2_train train.dataset_path=$DATASET_PATH
- Listen, Attend and Spell Training
python ./bin/main.py model=las train=las_train train.dataset_path=$DATASET_PATH
- Joint CTC-Attention Listen, Attend and Spell Training
python ./bin/main.py model=joint-ctc-attention-las train=las_train train.dataset_path=$DATASET_PATH
- Speech Transformer Training
python ./bin/main.py model=transformer train=transformer_train train.dataset_path=$DATASET_PATH
- Joint CTC-Attention Speech Transformer Training
python ./bin/main.py model=joint-ctc-attention-transformer train=transformer_train train.dataset_path=$DATASET_PATH
python ./bin/eval.py dataset_path=$DATASET_PATH transcripts_path=$TRANSCRIPTS_PATH model_path=$MODEL_PATH
Now you have a model which you can use to predict on new data. We do this by running greedy search
or beam search
.
- Command
$ python3 inference.py --model_path $MODEL_PATH --audio_path $AUDIO_PATH --device $DEVICE
- Output
음성인식 결과 문장이 나옵니다
You can get a quick look of pre-trained model's inference, with a audio.
Checkpoints are organized by experiments and timestamps as shown in the following file structure.
outputs
+-- YYYY_mm_dd
| +-- HH_MM_SS
| +-- trainer_states.pt
| +-- model.pt
You can resume and load from checkpoints.
If you have any questions, bug reports, and feature requests, please open an issue on Github.
For live discussions, please go to our gitter or Contacts [email protected] please.
We appreciate any kind of feedback or contribution. Feel free to proceed with small issues like bug fixes, documentation improvement. For major contributions and new features, please discuss with the collaborators in corresponding issues.
We follow PEP-8 for code style. Especially the style of docstrings is important to generate documentation.
Ilya Sutskever et al. Sequence to Sequence Learning with Neural Networks arXiv: 1409.3215
Dzmitry Bahdanau et al. Neural Machine Translation by Jointly Learning to Align and Translate arXiv: 1409.0473
Jan Chorowski et al. Attention Based Models for Speech Recognition arXiv: 1506.07503
Wiliam Chan et al. Listen, Attend and Spell arXiv: 1508.01211
Dario Amodei et al. Deep Speech2: End-to-End Speech Recognition in English and Mandarin arXiv: 1512.02595
Takaaki Hori et al. Advances in Joint CTC-Attention based E2E Automatic Speech Recognition with a Deep CNN Encoder and RNN-LM arXiv: 1706.02737
Ashish Vaswani et al. Attention Is All You Need arXiv: 1706.03762
Chung-Cheng Chiu et al. State-of-the-art Speech Recognition with Sequence-to-Sequence Models arXiv: 1712.01769
Anjuli Kannan et al. An Analysis Of Incorporating An External LM Into A Sequence-to-Sequence Model arXiv: 1712.01996
Daniel S. Park et al. SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition arXiv: 1904.08779
Rafael Muller et al. When Does Label Smoothing Help? arXiv: 1906.02629
Daniel S. Park et al. SpecAugment on large scale datasets arXiv:1912.05533
Jung-Woo Ha et al. ClovaCall: Korean Goal-Oriented Dialog Speech Corpus for Automatic Speech Recognition of Contact Centers arXiv: 2004.09367
AppleHolic/2020 AI Challenge - SpeechRecognition
This project is licensed under the Apache-2.0 LICENSE - see the LICENSE.md file for details
A technical report on KoSpeech is available. If you use the system for academic work, please cite:
@ARTICLE{2020kospeech,
author = {Soohwan Kim, Seyoung Bae, Cheolhwang Won},
title = "{KoSpeech: Open-Source Toolkit for End-to-End Korean Speech Recognition}",
journal = {ArXiv e-prints},
eprint = {2009.03092}
}