09 Sep 2023 2585字 9分
CC BY 4.0 （除特别声明或转载文章外）
如果这篇博客帮助到你，可以请我喝一杯咖啡~

Audio course

合在一起

语音到语音翻译：将语音从一种语言翻译成不同语言的语音
创建语音助手：建立自己的语音助手，该助手以与Alexa或Siri相似的方式工作
抄录会议：抄录会议并将笔录标记与谁在

语音到语音翻译

STST or S2ST

speech translation (ST) system

用三步达到目的

之前一般是 ASR + MT+ TTS

Speech translation

whisper 模型

import torch
from transformers import pipeline

device = "cuda:0" if torch.cuda.is_available() else "cpu"
pipe = pipeline(
    "automatic-speech-recognition", model="openai/whisper-base", device=device
)

用意大利语测试 VoxPopuli 数据集

from datasets import load_dataset

dataset = load_dataset("facebook/voxpopuli", "it", split="validation", streaming=True)
sample = next(iter(dataset))

看一下

from IPython.display import Audio

Audio(sample["audio"]["array"], rate=sample["audio"]["sampling_rate"])

定义函数

def translate(audio):
    outputs = pipe(audio, max_new_tokens=256, generate_kwargs={"task": "translate"})
    return outputs["text"]

再看一下

translate(sample["audio"].copy())

对比

sample["raw_text"]

Text to speech

加载 SpeechT5

from transformers import SpeechT5Processor, SpeechT5ForTextToSpeech, SpeechT5HifiGan

processor = SpeechT5Processor.from_pretrained("microsoft/speecht5_tts")

model = SpeechT5ForTextToSpeech.from_pretrained("microsoft/speecht5_tts")
vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan")

model.to(device)
vocoder.to(device)

embeddings_dataset = load_dataset("Matthijs/cmu-arctic-xvectors", split="validation")
speaker_embeddings = torch.tensor(embeddings_dataset[7306]["xvector"]).unsqueeze(0)

def synthesise(text):
    inputs = processor(text=text, return_tensors="pt")
    speech = model.generate_speech(
        inputs["input_ids"].to(device), speaker_embeddings.to(device), vocoder=vocoder
    )
    return speech.cpu()

查看一下

speech = synthesise("Hey there! This is a test!")

Audio(speech, rate=16000)

创建 STST demo

import numpy as np

target_dtype = np.int16
max_range = np.iinfo(target_dtype).max


def speech_to_speech_translation(audio):
    translated_text = translate(audio)
    synthesised_speech = synthesise(translated_text)
    synthesised_speech = (synthesised_speech.numpy() * max_range).astype(np.int16)
    return 16000, synthesised_speech

sampling_rate, synthesised_speech = speech_to_speech_translation(sample["audio"])

Audio(synthesised_speech, rate=sampling_rate)

import gradio as gr

demo = gr.Blocks()

mic_translate = gr.Interface(
    fn=speech_to_speech_translation,
    inputs=gr.Audio(source="microphone", type="filepath"),
    outputs=gr.Audio(label="Generated Speech", type="numpy"),
)

file_translate = gr.Interface(
    fn=speech_to_speech_translation,
    inputs=gr.Audio(source="upload", type="filepath"),
    outputs=gr.Audio(label="Generated Speech", type="numpy"),
)

with demo:
    gr.TabbedInterface([mic_translate, file_translate], ["Microphone", "Audio File"])

demo.launch(debug=True)

创建语音助手

端到端的 Marvin 来了

Wake word detection(唤醒词)
Speech transcription（语音转写）
Language model query（语言模型查询）
Synthesise speech（语音合成）

Audio Course 7