Hugging Face's logo

Spaces:
Detomo
/
ai-avatar-backend
Running

App Files Community
ai-avatar-backend
File size: 2,780 Bytes 
14afd56
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
 
// azure-cognitiveservices-speech.js
require('dotenv').config()
const sdk = require('microsoft-cognitiveservices-speech-sdk');
const blendShapeNames = require('./blendshapeNames');
const _ = require('lodash');
const voicesMap = {
    'en-US': 'en-US-AmberNeural',
    'ja-JP': 'ja-JP-MayuNeural',
    'vi-VN': 'vi-VN-NamMinhNeural',
};

let SSML = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="http://www.w3.org/2001/mstts" xml:lang="en-US">
<voice name="en-US-JennyNeural">
  <mstts:viseme type="FacialExpression"/>
  __TEXT__
</voice>
</speak>`;

const key = process.env.AZURE_KEY;
const region = process.env.AZURE_REGION;
        
/**
 * Node.js server code to convert text to speech
 * @returns stream
 * @param {*} text text to convert to audio/speech
 * @param language
 */
const textToSpeech = async (text, language)=> {
    
    // convert callback function to promise
    return new Promise((resolve, reject) => {

        const voice = voicesMap[language];
        let ssml = SSML.replace("__TEXT__", text).replace("<voice name=\"en-US-JennyNeural\">", `<voice name="${voice}">`);

        
        const speechConfig = sdk.SpeechConfig.fromSubscription(key, region);
        speechConfig.speechSynthesisOutputFormat = 5; // mp3
        
        let audioConfig = null;
        
        // if (filename) {
        let randomString = Math.random().toString(36).slice(2, 7);
        let filename = `./public/speech-${randomString}.mp3`;
        audioConfig = sdk.AudioConfig.fromAudioFileOutput(filename);
        // }

        let blendData = [];
        let timeStep = 1/60;
        let timeStamp = 0;

        const synthesizer = new sdk.SpeechSynthesizer(speechConfig, audioConfig);

        // Subscribes to viseme received event
        synthesizer.visemeReceived = function (s, e) {

            // `Animation` is an xml string for SVG or a json string for blend shapes
            var animation = JSON.parse(e.animation);

            _.each(animation.BlendShapes, blendArray => {

                let blend = {};
                _.each(blendShapeNames, (shapeName, i) => {
                    blend[shapeName] = blendArray[i];
                });
        
                blendData.push({
                    time: timeStamp,
                    blendshapes: blend
                });
                timeStamp += timeStep;
            });

        }


        synthesizer.speakSsmlAsync(
            ssml,
            result => {
                
                synthesizer.close();
                resolve({blendData, filename: `/speech-${randomString}.mp3`});

            },
            error => {
                synthesizer.close();
                reject(error);
            }); 
    });
};

module.exports = textToSpeech;