GitHub - espnet/espnet: End-to-End Speech Processing Toolkit

End-to-end speech processing toolkit

Documentation · Installation · Recipes · Model Zoo · Notebooks · Discord

ESPnet is an end-to-end speech processing toolkit built on PyTorch. It covers speech recognition, text-to-speech, speech translation, speech enhancement, speaker diarization, spoken language understanding, singing voice synthesis, speech language models, and more — with Kaldi-style reproducible recipes from data preparation to evaluation, and hundreds of pretrained models on Hugging Face.

What's new

ESPnet 202610.post1 — the command line grows espnet demo (the OWSM browser demo) and espnet asr --live (the microphone, transcribed as you speak); one Speech2Text now loads either kind of OWSM checkpoint, with best_path() for CTC decoding without a search; espnet/espnet:inference-cpu-latest and -gpu-latest run a published model with nothing installed; three more demo Spaces (TTS, enhancement, speaker verification).

Earlier releases

ESPnet 202610 — one-line inference from the command line (pip install espnet && espnet asr audio.wav), two OWSM v4 demos as Hugging Face Spaces, a core install without the training stack (training is espnet[train]), batched beam search, PyTorch 2.11-2.14.
ESPnet 202609 — ESPnet3 complete on egs3/librispeech_100 at ESPnet2 parity, CI rebuilt on a prebuilt image (compute per run halved), OpenBEATs pretraining, ten new recipes (ASR, TTS, SER, ST, audio SSL), Python 3.12-3.13.
ESPnet 202604 — Docker-based CI, PyTorch 2.9.1 support, FastSpeech2 inference ~1.9x faster at batch 8, new recipes (Kinyarwanda, Emilia, kosp2e).
ESPnet 202511 — parallel-processing primitives, refactored inference and evaluation pipeline, expanded SpeechLM support.
ESPnet 202509 — Python 3.9-3.13, Debian 12 CI, the LID subsystem completed, multi-optimizer training (HybridOptim / HybridLRS).
ESPnet 202506 — ESPnet3 groundwork (data organizer, trainer, model), LID training and task setup, codec1 recipes, USES2 speech enhancement, IPAPack++ S2T recipes.
ESPnet 202503 — PyTorch Lightning trainer support, Hugging Face front-end, scaled dot-product attention, ML-SUPERB 2024 recipe.

Full history: Releases.

Install

# Install PyTorch first: https://pytorch.org/get-started/locally/
pip install espnet              # run pretrained models
pip install "espnet[train]"     # also train them, with espnet2 or espnet3 (Lightning, TensorBoard, W&B, Hydra, Dask, ...)

Other installation options

pip install "espnet[all]"                       # training plus every task extra (except sds)
pip install git+https://github.com/espnet/espnet  # latest master

Full setup (recipes, DNN training, Kaldi-style tooling): see the installation guide.
Docker: see docker/ and the Docker docs.
Task-specific tools live in tools/installers.
ESPnet1 is no longer supported — use ESPnet2 (egs2/) or ESPnet3 (egs3/). See the ESPnet1 notice.

Tested environments (CI status)

system/pytorch ver.	2.11.0	2.13.0	2.14.0
ubuntu/python3.12/pip
ubuntu/python3.13/pip
debian12/python3.12/conda
windows/python3.12/pip
macos/python3.12/pip
macos/python3.12/conda

Each badge is its workflow's aggregate status on master, where the full grid runs. Coverage is not uniform — some suites run on one pytorch only, and a pull request runs less than master does. What each column covers.

Quick start

From the terminal — no code, on any audio file:

pip install espnet
espnet asr audio.wav                       # transcribe, detecting the language
espnet translate audio.wav --to eng        # speech in, English text out
espnet tts "Hello from ESPnet" -o out.wav
espnet enhance noisy.wav -o clean.wav
espnet asr --live                          # transcribe the microphone
espnet models                              # the default model of each command
pip install "espnet[demo]"
espnet demo                                # the same model in your browser

Every command takes --model <tag> and --device cuda, and espnet --version names the installed version. espnet demo serves, on localhost, the app behind the OWSM-CTC v4 Space — recording, upload, the checkpoint's own language and translation menus, and long-form decoding — and takes --port and --share.

Without installing anything — the same commands, in a container:

docker run --rm -v "$PWD:/data" -v "$HOME/.cache/huggingface:/cache/huggingface" \
    espnet/espnet:inference-cpu-latest asr /data/audio.wav

With a GPU, espnet/espnet:inference-gpu-latest, --gpus all and --device cuda.

The second mount is what keeps the downloaded model between runs; on a Linux host add --user "$(id -u):$(id -g)", so that what it writes belongs to you. The other two images, and what each is for, are in docker/.

From Python — any model from the ESPnet Hugging Face organization:

from espnet2.bin.s2t_inference import Speech2Text

# OWSM-CTC v4: multilingual ASR, translation and language ID in one
# encoder-only model. No beam search: one encoder pass per 30 s window.
s2t = Speech2Text.from_pretrained(
    "espnet/owsm_ctc_v4_1B", lang_sym="<eng>", task_sym="<asr>"
)
for start, end, text in s2t.decode_long("audio.wav"):  # any length or rate
    print(text)

The 4 GB checkpoint is cached after the first download. device="cuda" runs on a GPU, task_sym="<st_deu>" translates, lang_sym="<nolang>" identifies the language. The same class loads the encoder-decoder OWSM v4 models: decode_long then decodes segment by segment on the model's own timestamps, s2t(speech) runs the beam search, and s2t.best_path(speech) is the CTC head with no search. Every other task — asr_inference, tts_inference, enh_inference, spk_inference — follows the same from_pretrained pattern.

From an agent — pip install "espnet[mcp]", then register espnet-mcp as an MCP server (Claude Code: claude mcp add espnet -- espnet-mcp). Agents then call transcribe, synthesize and enhance themselves.

Train a recipe — every corpus follows the same interface:

cd egs2/librispeech/asr1
./run.sh                              # data → features → training → scoring
./run.sh --stage 11 --stop_stage 13   # or selected stages

New to ESPnet? Start with egs2/mini_an4/asr1 — it runs end to end in minutes.

Supported tasks

	Task	Template	Highlights
🗣️	ASR — speech recognition	`asr1`, `asr2`	Hybrid CTC/attention, Transducer, streaming, Conformer / E-Branchformer, Whisper, SSL front-ends
🌏	S2T — multilingual multitask	`s2t1`	OWSM: open Whisper-style models trained on public data
🔊	TTS — text-to-speech	`tts1`, `tts2`	Tacotron 2, FastSpeech 2, VITS, JETS, multi-speaker / multilingual
🎤	SVS — singing voice synthesis	`svs1`, `svs2`	VISinger 1/2, Xiaoice, DiffSinger; merged from Muskits
🎧	SE/SS — enhancement & separation	`enh1`, `enh_asr1`	Unified encoder–separator–decoder, TasNet / DPRNN / beamformers, ASR-integrated
🌐	ST / MT / S2ST — translation	`st1`, `mt1`, `s2st1`	End-to-end and cascaded speech translation, speech-to-speech translation
💬	SLU — language understanding	`slu1`	Intent + transcript multitasking, pretrained ASR/NLP encoders
👤	SPK / LID / DIAR — speaker & language	`spk1`, `lid1`, `diar1`	Speaker embeddings, verification, language ID, diarization
🧠	SSL — self-supervised learning	`ssl1`, `hubert1`	HuBERT pretraining; S3PRL upstreams as front-ends
🤖	SpeechLM — speech language models	`speechlm1`	Unified sequence modeling across speech and text tasks
📦	Codec — neural audio codecs	`codec1`	Discrete speech tokens for downstream tasks
➕	More	`uasr1`, `cls1`, `asvspoof1`, `lm1`, `sds1`	Unsupervised ASR (EURO), audio classification, anti-spoofing, LM, spoken dialogue

Each template ships a corpus-agnostic pipeline; see egs2/README.md for the full list of 200+ corpora recipes.

Why ESPnet

Reproducible — one run.sh per corpus, from download to scoring, with published results.
Unified — the same recipe structure, config format, and trainer across every task above.
Scalable — DDP, multi-node training, Slurm/MPI, DeepSpeed, sharded training, on-the-fly feature extraction.
Open — hundreds of pretrained models and demos on Hugging Face, plus W&B and TensorBoard logging.

Demos

Demo		Last run
Spoken dialogue — listen, think, speak
Speech recognition, in any of 151 languages
The words appearing as the audio arrives
Speech translation — the same model, a different task symbol
Text-to-speech, one voice and then 128
Speech enhancement, and what it did to the signal-to-noise
Speaker verification — two recordings, one score
Neural codecs — a waveform as a few integers a frame

Each runs top to bottom on a CPU and pins the release it was checked against. The second badge is that notebook being executed cell by cell every Sunday, so a red one names the demo that broke rather than leaving you to find out by opening it. More, including the CMU course material: espnet/notebook.

The full spoken dialogue system — microphone, voice activity detection, an end-to-end option beside the cascade, and latency and quality measured while you talk — is egs2/TEMPLATE/sds1, which runs locally.

Publish your own. Every ESPnet3 recipe can wrap its trained model in a Gradio app and push it to Hugging Face Spaces — the UI, the Space README.md and requirements.txt are all generated from conf/demo.yaml.

The three stages

cd egs3/librispeech_100/asr
train=conf/tuning/training_e_branchformer.yaml   # the config the model was trained with
python run.py --stages pack_model  --training_config $train --publication_config conf/publication.yaml  # -> exp/.../model_pack
python run.py --stages pack_demo   --training_config $train --demo_config conf/demo.yaml                # -> demo/
python run.py --stages upload_demo --training_config $train --demo_config conf/demo.yaml                # needs `hf auth login`

Run the packed app locally with python demo/app.py.

Learn

Documentation · ESPnet2 tutorial
Course tutorials at CMU: usage · adding new models/tasks (materials)
Interspeech 2019 tutorial

Contributing

Contributions, questions, and feature requests are all welcome — open an issue or a pull request. First time here? Read the contribution guide.

Details

Full feature list by task

Kaldi-style complete recipe

Support numbers of ASR recipes (WSJ, Switchboard, CHiME-4/5, Librispeech, TED, CSJ, AMI, HKUST, Voxforge, REVERB, Gigaspeech, etc.)
Support numbers of TTS recipes in a similar manner to the ASR recipe (LJSpeech, LibriTTS, M-AILABS, etc.)
Support numbers of ST recipes (Fisher-CallHome Spanish, Libri-trans, IWSLT'18, How2, Must-C, Mboshi-French, etc.)
Support numbers of MT recipes (IWSLT'14, IWSLT'16, the above ST recipes etc.)
Support numbers of SLU recipes (CATSLU-MAPS, FSC, Grabo, IEMOCAP, JDCINAL, SNIPS, SLURP, SWBD-DA, etc.)
Support numbers of SE/SS recipes (DNS-IS2020, LibriMix, SMS-WSJ, VCTK-noisyreverb, WHAM!, WHAMR!, WSJ-2mix, etc.)
Support voice conversion recipe (VCC2020 baseline)
Support speaker diarization recipe (mini_librispeech, librimix)
Support singing voice synthesis recipe (ofuton_p_utagoe_db, opencpop, m4singer, etc.)

ASR: Automatic Speech Recognition

State-of-the-art performance in several ASR benchmarks (comparable/superior to hybrid DNN/HMM and CTC)
Hybrid CTC/attention based end-to-end ASR
- Fast/accurate training with CTC/attention multitask training
- CTC/attention joint decoding to boost monotonic alignment decoding
- Encoder: VGG-like CNN + BiRNN (LSTM/GRU), sub-sampling BiRNN (LSTM/GRU), Transformer, Conformer, Branchformer, or E-Branchformer
- Decoder: RNN (LSTM/GRU), Transformer, or S4
Attention: Flash Attention, Dot product, location-aware attention, variants of multi-head
Incorporate RNNLM/LSTMLM/TransformerLM/N-gram trained only with text data
Batch GPU decoding
Data augmentation
Transducer based end-to-end ASR
- Architecture:
  - Custom encoder supporting RNNs, Conformer, Branchformer (w/ variants), 1D Conv / TDNN.
  - Decoder w/ parameters shared across blocks supporting RNN, stateless w/ 1D Conv, MEGA, and RWKV.
  - Pre-encoder: VGG2L or Conv2D available.
- Search algorithms:
  - Greedy search constrained to one emission by timestep.
  - Default beam search algorithm [Graves, 2012] without prefix search.
  - Alignment-Length Synchronous decoding [Saon et al., 2020].
  - Time Synchronous Decoding [Saon et al., 2020].
  - N-step Constrained beam search modified from [Kim et al., 2020].
  - modified Adaptive Expansion Search based on [Kim et al., 2021] and NSC.
- Features:
  - Unified interface for offline and streaming speech recognition.
  - Multi-task learning with various auxiliary losses:
    - Encoder: CTC, auxiliary Transducer and symmetric KL divergence.
    - Decoder: cross-entropy w/ label smoothing.
  - Transfer learning with an acoustic model and/or language model.
  - Training with FastEmit regularization method [Yu et al., 2021].
Please refer to the tutorial page for complete documentation.
CTC segmentation
Non-autoregressive model based on Mask-CTC
ASR examples for supporting endangered language documentation (see egs2/puebla_nahuatl and egs2/yoloxochitl_mixtec)
Wav2Vec2.0 pre-trained model as Encoder, imported from FairSeq.
Self-supervised learning representations as features, using upstream models in S3PRL in frontend.
- Set frontend to s3prl
- Select any upstream model by setting the frontend_conf to the corresponding name.
Transfer Learning :
- easy usage and transfers from models previously trained by your group or models from ESPnet Hugging Face repository.
- Documentation, and an old notebook that shows the idea. It is unmaintained and nothing runs it, so expect to fix it before it works.
Streaming Transformer/Conformer ASR with blockwise synchronous beam search.
Restricted Self-Attention based on Longformer as an encoder for long sequences
OpenAI Whisper model, robust ASR based on large-scale, weakly-supervised multitask learning

Demonstration

Speech recognition with OWSM-CTC
Hosted demo: OWSM-CTC v4, maintained from egs2/owsm_ctc_v4/s2t1/demo — recognises and identifies 151 languages, translates into 25 of them, and decodes long-form audio
Hosted demo: OWSM v4, maintained from egs2/owsm_v4/s2t1/demo — the same four tasks with the encoder-decoder model, which also takes a text prompt
Streaming ASR, decoded as the audio arrives

TTS: Text-to-speech

Architecture
- Tacotron2
- Transformer-TTS
- FastSpeech
- FastSpeech2
- Conformer FastSpeech & FastSpeech2
- VITS
- JETS
Multi-speaker & multi-language extension
- Pre-trained speaker embedding (e.g., X-vector)
- Speaker ID embedding
- Language ID embedding
- Global style token (GST) embedding
- Mix of the above embeddings
End-to-end training
- End-to-end text-to-wav model (e.g., VITS, JETS, etc.)
- Joint training of text2mel and vocoder
Various language support
- En / Jp / Zn / De / Ru / And more...
Integration with neural vocoders
- Parallel WaveGAN
- MelGAN
- Multi-band MelGAN
- HiFiGAN
- StyleMelGAN
- Mix of the above models

Demonstration

TTS demo with ESPnet2
Integrated to Hugging Face Spaces with Gradio. See demo:

To train the neural vocoder, please check the following repositories:

SE: Speech enhancement (and separation)

Single-speaker speech enhancement
Multi-speaker speech separation
Unified encoder-separator-decoder structure for time-domain and frequency-domain models
- Encoder/Decoder: STFT/iSTFT, Convolution/Transposed-Convolution
- Separators: BLSTM, Transformer, Conformer, TasNet, DPRNN, SkiM, SVoice, DC-CRN, DCCRN, Deep Clustering, Deep Attractor Network, FaSNet, iFaSNet, Neural Beamformers, etc.
Flexible ASR integration: working as an individual task or as the ASR frontend
Easy to import pre-trained models from Asteroid
- Both the pre-trained models from Asteroid and the specific configuration are supported.

Demonstration

Interactive SE demo with ESPnet2
Streaming SE demo with ESPnet2

ST: Speech Translation & MT: Machine Translation

State-of-the-art performance in several ST benchmarks (comparable/superior to cascaded ASR and MT)
Transformer-based end-to-end ST (new!)
Transformer-based end-to-end MT (new!)

VC: Voice conversion

Transformer and Tacotron2-based parallel VC using Mel spectrogram
End-to-end VC based on cascaded ASR+TTS (Baseline system for Voice Conversion Challenge 2020!)

SLU: Spoken Language Understanding

Architecture
- Transformer-based Encoder
- Conformer-based Encoder
- Branchformer based Encoder
- E-Branchformer based Encoder
- RNN based Decoder
- Transformer-based Decoder
Support Multitasking with ASR
- Predict both intent and ASR transcript
Support Multitasking with NLU
- Deliberation encoder based 2 pass model
Support using pre-trained ASR models
- Hubert
- Wav2vec2
- VQ-APC
- TERA and more ...
Support using pre-trained NLP models
- BERT
- MPNet And more...
Various language support
- En / Jp / Zn / Nl / And more...
Supports using context from previous utterances
Supports using other tasks like SE in a pipeline manner
Supports Two Pass SLU that combines audio and ASR transcript Demonstration
Performing noisy spoken language understanding using a speech enhancement model followed by a spoken language understanding model.
Performing two-pass spoken language understanding where the second pass model attends to both acoustic and semantic information.

SUM: Speech Summarization

End to End Speech Summarization Recipe for Instructional Videos using Restricted Self-Attention [Sharma et al., 2022]

SVS: Singing Voice Synthesis

Framework merge from Muskits
Architecture
- RNN-based non-autoregressive model
- Xiaoice
- Tacotron-singing
- DiffSinger (in progress)
- VISinger
- VISinger 2 (its variations with different vocoders-architecture)
Support multi-speaker & multilingual singing synthesis
- Speaker ID embedding
- Language ID embedding
Various language support
- Jp / En / Kr / Zh
Tight integration with neural vocoders (the same as TTS)

SSL: Self-supervised Learning

Support HuBERT Pre-training:
- Example recipe: egs2/LibriSpeech/ssl1

UASR: Unsupervised ASR (EURO: ESPnet Unsupervised Recognition - Open-source)

Architecture
- wav2vec-U (with different self-supervised models)
- wav2vec-U 2.0 (in progress)
Support PrefixBeamSearch and K2-based WFST decoding

S2T: Speech-to-text with Whisper-style multilingual multitask models

Reproduces Whisper-style training from scratch using public data: OWSM
Supports multiple tasks in a single model
- Multilingual speech recognition
- Any-to-any speech translation
- Language identification
- Utterance-level timestamp prediction (segmentation)

DNN Framework

Flexible network architecture on PyTorch
Flexible front-end processing thanks to omniio and HDF5 support
Tensorboard-based monitoring
DeepSpeed-based large-scale training

ESPnet2

See ESPnet2.

Independent from Kaldi/Chainer, unlike ESPnet1
On-the-fly feature extraction and text processing when training
Supporting DistributedDataParallel and DaraParallel both
Supporting multiple nodes training and integrated with Slurm or MPI
Supporting Sharded Training provided by fairscale
A template recipe that can be applied to all corpora
Possible to train any size of corpus without CPU memory error
ESPnet Model Zoo
Integrated with wandb

Benchmark results and command-line demos

ASR results

expand

We list the character error rate (CER) and word error rate (WER) of major ASR tasks.

Task	CER (%)	WER (%)	Pre-trained model
Aishell dev/test	4.1/4.4	N/A	link
CSJ eval1/eval2/eval3	4.5/3.3/3.6	N/A	link
GigaSpeech dev/test	N/A	10.6/10.5	link
HKUST dev	21.2	N/A	link
Librispeech dev_clean/dev_other/test_clean/test_other	0.6/1.5/0.6/1.4	1.7/3.4/1.8/3.6	link
Switchboard (eval2000) callhm/swbd	N/A	13.4/7.3	link
TEDLIUM2 dev/test	N/A	7.3/7.1	link
WSJ dev93/eval92	1.1/0.8	2.8/1.8	link

If you want to check the results of the other recipes, please check egs2/<name_of_recipe>/asr1/README.md.

SE results

expand

We list results from three different models on WSJ0-2mix, which is one the most widely used benchmark dataset for speech separation.

Model	STOI	SAR	SDR	SIR
TF Masking	0.89	11.40	10.24	18.04
Conv-Tasnet	0.95	16.62	15.94	25.90
DPRNN-Tasnet	0.96	18.82	18.29	28.92

MT results

expand

Task	BLEU	Pre-trained model
IWSLT'14 test2014 (De->En)	32.2	link

TTS results

expand

You can listen to the generated samples in the following URL.

ESPnet2 TTS generated samples

Note that in the generation, we use Griffin-Lim (wav/) and Parallel WaveGAN (wav_pwg/).

You can download pre-trained models via espnet_model_zoo.

You can download pre-trained vocoders via kan-bayashi/ParallelWaveGAN.

TTS demo

expand

You can try the real-time demo in Google Colab. Please access the notebook from the following button and enjoy the real-time synthesis!

TTS demo with ESPnet2

English, Japanese, and Mandarin models are available in the demo.

VC results

expand

Transformer and Tacotron2-based VC

You can listen to some samples on the demo webpage.

Cascade ASR+TTS as one of the baseline systems of VCC2020

The Voice Conversion Challenge 2020 (VCC2020) adopts ESPnet to build an end-to-end based baseline system. In VCC2020, the objective is intra/cross-lingual nonparallel VC. You can download converted samples of the cascade ASR+TTS baseline system here.

SLU results

expand

We list the performance on various SLU tasks and datasets using the metric reported in the original dataset paper

Task	Dataset	Metric	Result	Pre-trained Model
Intent Classification	SLURP	Acc	86.3	link
Intent Classification	FSC	Acc	99.6	link
Intent Classification	FSC Unseen Speaker Set	Acc	98.6	link
Intent Classification	FSC Unseen Utterance Set	Acc	86.4	link
Intent Classification	FSC Challenge Speaker Set	Acc	97.5	link
Intent Classification	FSC Challenge Utterance Set	Acc	78.5	link
Intent Classification	SNIPS	F1	91.7	link
Intent Classification	Grabo (Nl)	Acc	97.2	link
Intent Classification	CAT SLU MAP (Zn)	Acc	78.9	link
Intent Classification	Google Speech Commands	Acc	98.4	link
Slot Filling	SLURP	SLU-F1	71.9	link
Dialogue Act Classification	Switchboard	Acc	67.5	link
Dialogue Act Classification	Jdcinal (Jp)	Acc	67.4	link
Emotion Recognition	IEMOCAP	Acc	69.4	link
Emotion Recognition	swbd_sentiment	Macro F1	61.4	link
Emotion Recognition	slue_voxceleb	Macro F1	44.0	link

If you want to check the results of the other recipes, please check egs2/<name_of_recipe>/asr1/RESULTS.md.

CTC Segmentation demo

expand

CTC segmentation determines utterance segments within audio files. Aligned utterance segments constitute the labels of speech datasets.

As a demo, we align the start and end of utterances within the audio file ctc_align_test.wav. This can be done either directly from the Python command line or using the script espnet2/bin/asr_align.py.

From the Python command line interface:

# load a model with character tokens
from espnet_model_zoo.downloader import ModelDownloader
d = ModelDownloader(cachedir="./modelcache")
wsjmodel = d.download_and_unpack("kamo-naoyuki/wsj")
# load the example file included in the ESPnet repository
import soundfile
speech, rate = soundfile.read("./test_utils/ctc_align_test.wav")
# CTC segmentation
from espnet2.bin.asr_align import CTCSegmentation
aligner = CTCSegmentation( **wsjmodel , fs=rate )
text = """
utt1 THE SALE OF THE HOTELS
utt2 IS PART OF HOLIDAY'S STRATEGY
utt3 TO SELL OFF ASSETS
utt4 AND CONCENTRATE ON PROPERTY MANAGEMENT
"""
segments = aligner(speech, text)
print(segments)
# utt1 utt 0.26 1.73 -0.0154 THE SALE OF THE HOTELS
# utt2 utt 1.73 3.19 -0.7674 IS PART OF HOLIDAY'S STRATEGY
# utt3 utt 3.19 4.20 -0.7433 TO SELL OFF ASSETS
# utt4 utt 4.20 6.10 -0.4899 AND CONCENTRATE ON PROPERTY MANAGEMENT

Aligning also works with fragments of the text. For this, set the gratis_blank option that allows skipping unrelated audio sections without penalty. It's also possible to omit the utterance names at the beginning of each line by setting kaldi_style_text to False.

aligner.set_config( gratis_blank=True, kaldi_style_text=False )
text = ["SALE OF THE HOTELS", "PROPERTY MANAGEMENT"]
segments = aligner(speech, text)
print(segments)
# utt_0000 utt 0.37 1.72 -2.0651 SALE OF THE HOTELS
# utt_0001 utt 4.70 6.10 -5.0566 PROPERTY MANAGEMENT

The script espnet2/bin/asr_align.py uses a similar interface. To align utterances:

# ASR model and config files from pre-trained model (e.g., from cachedir):
asr_config=<path-to-model>/config.yaml
asr_model=<path-to-model>/valid.*best.pth
# prepare the text file
wav="test_utils/ctc_align_test.wav"
text="test_utils/ctc_align_text.txt"
cat << EOF > ${text}
utt1 THE SALE OF THE HOTELS
utt2 IS PART OF HOLIDAY'S STRATEGY
utt3 TO SELL OFF ASSETS
utt4 AND CONCENTRATE
utt5 ON PROPERTY MANAGEMENT
EOF
# obtain alignments:
python espnet2/bin/asr_align.py --asr_train_config ${asr_config} --asr_model_file ${asr_model} --audio ${wav} --text ${text}
# utt1 ctc_align_test 0.26 1.73 -0.0154 THE SALE OF THE HOTELS
# utt2 ctc_align_test 1.73 3.19 -0.7674 IS PART OF HOLIDAY'S STRATEGY
# utt3 ctc_align_test 3.19 4.20 -0.7433 TO SELL OFF ASSETS
# utt4 ctc_align_test 4.20 4.97 -0.6017 AND CONCENTRATE
# utt5 ctc_align_test 4.97 6.10 -0.3477 ON PROPERTY MANAGEMENT

The output of the script can be redirected to a segments file by adding the argument --output segments. Each line contains the file/utterance name, utterance start and end times in seconds, and a confidence score; optionally also the utterance text. The confidence score is a probability in log space that indicates how well the utterance was aligned. If needed, remove bad utterances:

min_confidence_score=-7
# here, we assume that the output was written to the file `segments`
awk -v ms=${min_confidence_score} '{ if ($5 > ms) {print} }' segments

See the module documentation for more information. It is recommended to use models with RNN-based encoders (such as BLSTMP) for aligning large audio files; rather than using Transformer models that have a high memory consumption on longer audio data. The sample rate of the audio must be consistent with that of the data used in training; adjust with sox if needed.

Also, we can use this tool to provide token-level segmentation information if we prepare a list of tokens instead of that of utterances in the text file. See the discussion in #4278 (comment).

Citation

If you use ESPnet in your research, please cite the main paper:

@inproceedings{watanabe2018espnet,
  author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson {Enrique Yalta Soplin} and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
  title={{ESPnet}: End-to-End Speech Processing Toolkit},
  year={2018},
  booktitle={Proceedings of Interspeech},
  pages={2207--2211},
  doi={10.21437/Interspeech.2018-1456},
  url={http://dx.doi.org/10.21437/Interspeech.2018-1456}
}

Task-specific papers (TTS, ST, SE, SLU, SVS, UASR, S2T, SUM, SPK)

@inproceedings{hayashi2020espnet,
  title={{Espnet-TTS}: Unified, reproducible, and integratable open source end-to-end text-to-speech toolkit},
  author={Hayashi, Tomoki and Yamamoto, Ryuichi and Inoue, Katsuki and Yoshimura, Takenori and Watanabe, Shinji and Toda, Tomoki and Takeda, Kazuya and Zhang, Yu and Tan, Xu},
  booktitle={Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  pages={7654--7658},
  year={2020},
  organization={IEEE}
}
@inproceedings{inaguma-etal-2020-espnet,
    title = "{ESP}net-{ST}: All-in-One Speech Translation Toolkit",
    author = "Inaguma, Hirofumi  and
      Kiyono, Shun  and
      Duh, Kevin  and
      Karita, Shigeki  and
      Yalta, Nelson  and
      Hayashi, Tomoki  and
      Watanabe, Shinji",
    booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: System Demonstrations",
    month = jul,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/2020.acl-demos.34",
    pages = "302--311",
}
@article{hayashi2021espnet2,
  title={{ESP}net2-{TTS}: Extending the edge of {TTS} research},
  author={Hayashi, Tomoki and Yamamoto, Ryuichi and Yoshimura, Takenori and Wu, Peter and Shi, Jiatong and Saeki, Takaaki and Ju, Yooncheol and Yasuda, Yusuke and Takamichi, Shinnosuke and Watanabe, Shinji},
  journal={arXiv preprint arXiv:2110.07840},
  year={2021}
}
@inproceedings{li2020espnet,
  title={{ESPnet-SE}: End-to-End Speech Enhancement and Separation Toolkit Designed for {ASR} Integration},
  author={Chenda Li and Jing Shi and Wangyou Zhang and Aswin Shanmugam Subramanian and Xuankai Chang and Naoyuki Kamo and Moto Hira and Tomoki Hayashi and Christoph Boeddeker and Zhuo Chen and Shinji Watanabe},
  booktitle={Proceedings of IEEE Spoken Language Technology Workshop (SLT)},
  pages={785--792},
  year={2021},
  organization={IEEE},
}
@inproceedings{arora2021espnet,
  title={{ESPnet-SLU}: Advancing Spoken Language Understanding through ESPnet},
  author={Arora, Siddhant and Dalmia, Siddharth and Denisov, Pavel and Chang, Xuankai and Ueda, Yushi and Peng, Yifan and Zhang, Yuekai and Kumar, Sujay and Ganesan, Karthik and Yan, Brian and others},
  booktitle={ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  pages={7167--7171},
  year={2022},
  organization={IEEE}
}
@inproceedings{shi2022muskits,
  author={Shi, Jiatong and Guo, Shuai and Qian, Tao and Huo, Nan and Hayashi, Tomoki and Wu, Yuning and Xu, Frank and Chang, Xuankai and Li, Huazhe and Wu, Peter and Watanabe, Shinji and Jin, Qin},
  title={{Muskits}: an End-to-End Music Processing Toolkit for Singing Voice Synthesis},
  year={2022},
  booktitle={Proceedings of Interspeech},
  pages={4277-4281},
  url={https://www.isca-speech.org/archive/pdfs/interspeech_2022/shi22d_interspeech.pdf}
}
@inproceedings{lu22c_interspeech,
  author={Yen-Ju Lu and Xuankai Chang and Chenda Li and Wangyou Zhang and Samuele Cornell and Zhaoheng Ni and Yoshiki Masuyama and Brian Yan and Robin Scheibler and Zhong-Qiu Wang and Yu Tsao and Yanmin Qian and Shinji Watanabe},
  title={{ESPnet-SE++: Speech Enhancement for Robust Speech Recognition, Translation, and Understanding}},
  year=2022,
  booktitle={Proc. Interspeech 2022},
  pages={5458--5462},
}
@inproceedings{gao2023euro,
  title={{EURO: ESP}net unsupervised {ASR} open-source toolkit},
  author={Gao, Dongji and Shi, Jiatong and Chuang, Shun-Po and Garcia, Leibny Paola and Lee, Hung-yi and Watanabe, Shinji and Khudanpur, Sanjeev},
  booktitle={ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  pages={1--5},
  year={2023},
  organization={IEEE}
}
@inproceedings{peng2023reproducing,
  title={Reproducing {W}hisper-style training using an open-source toolkit and publicly available data},
  author={Peng, Yifan and Tian, Jinchuan and Yan, Brian and Berrebbi, Dan and Chang, Xuankai and Li, Xinjian and Shi, Jiatong and Arora, Siddhant and Chen, William and Sharma, Roshan and others},
  booktitle={2023 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)},
  pages={1--8},
  year={2023},
  organization={IEEE}
}
@inproceedings{sharma2023espnet,
  title={ESPnet-{SUMM}: Introducing a novel large dataset, toolkit, and a cross-corpora evaluation of speech summarization systems},
  author={Sharma, Roshan and Chen, William and Kano, Takatomo and Sharma, Ruchira and Arora, Siddhant and Watanabe, Shinji and Ogawa, Atsunori and Delcroix, Marc and Singh, Rita and Raj, Bhiksha},
  booktitle={2023 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)},
  pages={1--8},
  year={2023},
  organization={IEEE}
}
@article{jung2024espnet,
  title={{ESPnet-SPK}: full pipeline speaker embedding toolkit with reproducible recipes, self-supervised front-ends, and off-the-shelf models},
  author={Jung, Jee-weon and Zhang, Wangyou and Shi, Jiatong and Aldeneh, Zakaria and Higuchi, Takuya and Theobald, Barry-John and Abdelaziz, Ahmed Hussen and Watanabe, Shinji},
  journal={Proc. Interspeech 2024},
  year={2024}
}
@inproceedings{yan-etal-2023-espnet,
    title = "{ESP}net-{ST}-v2: Multipurpose Spoken Language Translation Toolkit",
    author = "Yan, Brian  and
      Shi, Jiatong  and
      Tang, Yun  and
      Inaguma, Hirofumi  and
      Peng, Yifan  and
      Dalmia, Siddharth  and
      Pol{\'a}k, Peter  and
      Fernandes, Patrick  and
      Berrebbi, Dan  and
      Hayashi, Tomoki  and
      Zhang, Xiaohui  and
      Ni, Zhaoheng  and
      Hira, Moto  and
      Maiti, Soumi  and
      Pino, Juan  and
      Watanabe, Shinji",
    booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
    year = "2023",
    publisher = "Association for Computational Linguistics",
    pages = "400--411",
}
@inproceedings{someki2022espnet,
  title={{ESPnet-ONNX}: Bridging a gap between research and production},
  author={Someki, Masao and Higuchi, Yosuke and Hayashi, Tomoki and Watanabe, Shinji},
  booktitle={2022 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)},
  year={2022},
  organization={IEEE}
}
@inproceedings{someki2024espnet,
  title={{ESPnet-EZ}: Python-only {ESPnet} for easy fine-tuning and integration},
  author={Someki, Masao and Choi, Kwanghee and Arora, Siddhant and Chen, William and Cornell, Samuele and Han, Jionghao and Peng, Yifan and Shi, Jiatong and Srivastav, Vaibhav and Watanabe, Shinji},
  booktitle={2024 IEEE Spoken Language Technology Workshop (SLT)},
  year={2024},
  organization={IEEE}
}
@inproceedings{shi2024espnet,
  title={{ESPnet-Codec}: Comprehensive Training and Evaluation of Neural Codecs for Audio, Music, and Speech},
  author={Shi, Jiatong and Tian, Jinchuan and Wu, Yihan and Jung, Jee-weon and Yip, Jia Qi and Masuyama, Yoshiki and Chen, William and Wu, Yuning and Tang, Yuxun and Baali, Massa and Alharhi, Dareen and Zhang, Dong and Deng, Ruifan and Srivastava, Tejes and Wu, Haibin and Liu, Alexander H. and Raj, Bhiksha and Jin, Qin and Song, Ruihua and Watanabe, Shinji},
  booktitle={2024 IEEE Spoken Language Technology Workshop (SLT)},
  pages={562--569},
  year={2024},
  organization={IEEE}
}
@inproceedings{tian-etal-2025-espnet,
  title={{ESP}net-{S}peech{LM}: An Open Speech Language Model Toolkit},
  author={Tian, Jinchuan and Shi, Jiatong and Chen, William and Arora, Siddhant and Masuyama, Yoshiki and Maekaku, Takashi and Wu, Yihan and Peng, Junyi and Bharadwaj, Shikhar and Zhao, Yiwen and Cornell, Samuele and Peng, Yifan and Yue, Xiang and Yang, Chao-Han Huck and Neubig, Graham and Watanabe, Shinji},
  booktitle={Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (System Demonstrations)},
  pages={116--124},
  year={2025},
  publisher={Association for Computational Linguistics}
}
@inproceedings{arora-etal-2025-espnet,
  title={{ESP}net-{SDS}: Unified Toolkit and Demo for Spoken Dialogue Systems},
  author={Arora, Siddhant and Peng, Yifan and Shi, Jiatong and Tian, Jinchuan and Chen, William and Bharadwaj, Shikhar and Futami, Hayato and Kashiwagi, Yosuke and Tsunoo, Emiru and Shimizu, Shuichiro and Srivastav, Vaibhav and Watanabe, Shinji},
  booktitle={Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (System Demonstrations)},
  pages={248--259},
  year={2025},
  publisher={Association for Computational Linguistics}
}
@inproceedings{bharadwaj2025openbeats,
  title={{OpenBEATs}: A Fully Open-Source General-Purpose Audio Encoder},
  author={Bharadwaj, Shikhar and Cornell, Samuele and Choi, Kwanghee and Fukayama, Satoru and Shim, Hye-jin and Deshmukh, Soham and Watanabe, Shinji},
  booktitle={2025 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)},
  year={2025},
  organization={IEEE}
}
@inproceedings{someki2026espnet3,
  title={{ESPnet3}: Infrastructure for Scalable Speech and Audio Research in the Foundation Model Era},
  author={Someki, Masao and Polok, Alexander and Carvalho, Carlos and Lin, Chyi-Jiunn and Yang, Da-Hee and Shi, Jiatong and Tian, Jinchuan and Yalta Soplin, Nelson Enrique and Cornell, Samuele and Arora, Siddhant and Teixeira, Francisco and Wang, Wei and Chen, William and Abad, Alberto and Li, Chenda and Watanabe, Shinji and Zhang, Wangyou},
  booktitle={Interspeech 2026},
  year={2026}
}

Released under the Apache 2.0 License.

Name		Name	Last commit message	Last commit date
.devcontainer		.devcontainer
.github		.github
ci		ci
doc		doc
docker		docker
egs2		egs2
egs3		egs3
espnet		espnet
espnet2		espnet2
espnet3		espnet3
test		test
test_utils		test_utils
tools		tools
utils		utils
.coderabbit.yaml		.coderabbit.yaml
.continueignore		.continueignore
.coveragerc		.coveragerc
.dockerignore		.dockerignore
.gitignore		.gitignore
.gitmodules		.gitmodules
.mergify.yml		.mergify.yml
.pre-commit-config.yaml		.pre-commit-config.yaml
CONTRIBUTING.md		CONTRIBUTING.md
LICENSE		LICENSE
README.md		README.md
codecov.yml		codecov.yml
pyproject.toml		pyproject.toml
setup.cfg		setup.cfg
version.txt		version.txt

Latest commit

History

Folders and files

Repository files navigation

End-to-end speech processing toolkit

What's new

Install

Quick start

Supported tasks

Why ESPnet

Demos

Learn

Contributing

Details

Kaldi-style complete recipe

ASR: Automatic Speech Recognition

TTS: Text-to-speech

SE: Speech enhancement (and separation)

ST: Speech Translation & MT: Machine Translation

VC: Voice conversion

SLU: Spoken Language Understanding

SUM: Speech Summarization

SVS: Singing Voice Synthesis

SSL: Self-supervised Learning

UASR: Unsupervised ASR (EURO: ESPnet Unsupervised Recognition - Open-source)

S2T: Speech-to-text with Whisper-style multilingual multitask models

DNN Framework

ESPnet2

ASR results

SE results

MT results

TTS results

TTS demo

VC results

SLU results

CTC Segmentation demo

Citation

About

Topics

Resources

Contributing

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages