reazon_speech2

ReazonSpeech v2.0

Input

Audio file

speech-001.mov

Output

Recognized speech text

[00:00:00.300 --> 00:00:04.780] 気象庁は雪や路面の凍結による交通への影響、
[00:00:05.179 --> 00:00:12.620] 暴風雪や高波に警戒するとともに雪崩や屋根からの落雪にも十分注意するよう呼びかけています。

Requirements

This model requires additional module.

pip3 install librosa
pip3 install sentencepiece

Usage

Automatically downloads the onnx and prototxt files on the first run. It is necessary to be connected to the Internet while downloading.

For the sample wav,

$ python3 reazon_speech2.py

If you want to specify the audio, put the file path after the --input option.

$ python3 reazon_speech2.py --input AUDIO_FILE

Reference

• ReazonSpeech

Framework

Pytorch

Model Format

ONNX opset=17

Netron

reazonspeech-nemo-v2_encoder.onnx.prototxt
reazonspeech-nemo-v2_decoder.onnx.prototxt
reazonspeech-nemo-v2_joint.onnx.prototxt