检测到静音 (JS) 时，如何提取前面的音频(从麦克风)作为缓冲区?

我正在使用Google Cloud API进行语音转文本，并带有NodeJS后端。应用需要能够侦听语音命令，并将其作为缓冲区传输到后端。为此，我需要在检测到静音时发送前面音频的缓冲区。

任何帮助将不胜感激。包括下面的 js 代码

if (!navigator.getUserMedia)
navigator.getUserMedia = navigator.getUserMedia || navigator.webkitGetUserMedia ||
navigator.mozGetUserMedia || navigator.msGetUserMedia;
if (navigator.getUserMedia) {
navigator.getUserMedia({audio: true}, success, function (e) {
alert('Error capturing audio.');
});
} else alert('getUserMedia not supported in this browser.');
var recording = false;
window.startRecording = function () {
recording = true;
};
window.stopRecording = function () {
recording = false;
// window.Stream.end();
};
function success(e) {
audioContext = window.AudioContext || window.webkitAudioContext;
context = new audioContext();
// the sample rate is in context.sampleRate
audioInput = context.createMediaStreamSource(e);
var bufferSize = 4096;
recorder = context.createScriptProcessor(bufferSize, 1, 1);
recorder.onaudioprocess = function (e) {
if (!recording) return;
console.log('recording');
var left = e.inputBuffer.getChannelData(0);
console.log(convertoFloat32ToInt16(left));
};
audioInput.connect(recorder);
recorder.connect(context.destination);
}

我不太确定问题中到底问了什么，所以这个答案只是为了提供一种检测 AudioStream 中静音的方法。

要检测音频流中的静音，可以使用音频分析器节点，在该节点上定期调用getByteFrequencyData方法，并检查在给定时间内是否有高于预期电平的声音。

您可以使用分析器节点的minDecibels属性直接设置阈值水平。

function detectSilence(
stream,
onSoundEnd = _=>{},
onSoundStart = _=>{},
silence_delay = 500,
min_decibels = -80
) {
const ctx = new AudioContext();
const analyser = ctx.createAnalyser();
const streamNode = ctx.createMediaStreamSource(stream);
streamNode.connect(analyser);
analyser.minDecibels = min_decibels;
const data = new Uint8Array(analyser.frequencyBinCount); // will hold our data
let silence_start = performance.now();
let triggered = false; // trigger only once per silence event
function loop(time) {
requestAnimationFrame(loop); // we'll loop every 60th of a second to check
analyser.getByteFrequencyData(data); // get current data
if (data.some(v => v)) { // if there is data above the given db limit
if(triggered){
triggered = false;
onSoundStart();
}
silence_start = time; // set it to now
}
if (!triggered && time - silence_start > silence_delay) {
onSoundEnd();
triggered = true;
}
}
loop();
}
function onSilence() {
console.log('silence');
}
function onSpeak() {
console.log('speaking');
}
navigator.mediaDevices.getUserMedia({
audio: true
})
.then(stream => {
detectSilence(stream, onSilence, onSpeak);
// do something else with the stream
})
.catch(console.error);

并且作为一个小提琴，因为stackSnippets可能会阻止gUM。

您可以使用SpeechRecognitionresult事件来确定何时识别单词或短语，例如，ls、cd、pwd或其他命令，将SpeechRecognitionAlternative的.transcript传递到speechSynthesis.speak()在附加start和end事件SpeechSynthesisUtterance调用传递MediaStream的对象MediaRecorder上的.start()或.resume();将Blobdataavailable事件转换为使用FileReader或Response.arrayBuffer()的ArrayBuffer。

或者，我们可以将audiostart或soundstart与audioend或soundendSpeechRecognition事件一起使用来记录用户的实际声音，尽管与仅由标准系统麦克风捕获的音频的实际开始和结束可能不会一致地触发结束。

<!DOCTYPE html>
<html>
<head>
<title>Speech Recognition Recording</title>
</head>
<body>
<input type="button" value="Stop speech command recognition" id="stop">
<script>
navigator.mediaDevices.getUserMedia({
audio: true
})
.then(stream => {
const recorder = new MediaRecorder(stream);
const recognition = new webkitSpeechRecognition();
const synthesis = new SpeechSynthesisUtterance();
const handleResult = e => {
recognition.onresult = null;
console.log(e.results);
const result = e.results[e.results.length - 1];
if (result.isFinal) {
const [{transcript}] = result;
console.log(transcript);
synthesis.text = transcript;
window.speechSynthesis.speak(synthesis);
}
}
synthesis.onstart = () => {
if (recorder.state === "inactive") {
recorder.start()
} else {
if (recorder.state === "paused") {
recorder.resume();
}
}
}
synthesis.onend = () => {
recorder.pause();
recorder.requestData();
}
recorder.ondataavailable = async(e) => {
if (stream.active) {
try {
const blobURL = URL.createObjectURL(e.data);
const request = await fetch(blobURL);
const ab = await request.arrayBuffer();
console.log(blobURL, ab);
recognition.onresult = handleResult;
// URL.revokeObjectURL(blobURL);
} catch (err) {
throw err
}
}
}
recorder.onpause = e => {
console.log("recorder " + recorder.state);
}
recognition.continuous = true;
recognition.interimResults = false;
recognition.maxAlternatives = 1;
recognition.start();
recognition.onend = e => {
console.log("recognition ended, stream.active", stream.active);
if (stream.active) {
console.log(e);
// the service disconnects after a period of time
recognition.start();
}
}
recognition.onresult = handleResult;
stream.oninactive = () => {
console.log("stream ended");
}
document.getElementById("stop")
.onclick = () => {
console.log("stream.active:", stream.active);
if (stream && stream.active && recognition) {
recognition.abort();
recorder.stop();
for (let track of stream.getTracks()) {
track.stop();
}
console.log("stream.active:", stream.active);
}
}
})
.catch(err => {
console.error(err)
});
</script>
</body>
</html>

普罗提 https://plnkr.co/edit/4DVEg6mhFRR94M5gdaIp?p=preview

最简单的方法是使用.pause()和.resume()，.stop()MediaRecorder()的方法允许用户开始，暂停和停止录制使用navigator.mediaDevices.getUserMedia()捕获的音频并将生成的Blob转换为ArrayBuffer，如果这是API期望POST到服务器

的内容

<!DOCTYPE html>
<html>
<head>
<title>User Media Recording</title>
</head>
<body>
<input type="button" value="Start/resume recording audio" id="start">
<input type="button" value="Pause recording audio" id="pause">
<input type="button" value="Stop recording audio" id="stop">
<script>
navigator.mediaDevices.getUserMedia({
audio: true
})
.then(stream => {
const recorder = new MediaRecorder(stream);
recorder.ondataavailable = async(e) => {
if (stream.active) {
try {
const blobURL = URL.createObjectURL(e.data);
const request = await fetch(blobURL);
const ab = await request.arrayBuffer();
// do stuff with `ArrayBuffer` of recorded audio
console.log(blobURL, ab);
// we do not need the `Blob URL`, we can revoke the object
// URL.revokeObjectURL(blobURL);
} catch (err) {
throw err
}
}
}
recorder.onpause = e => {
console.log("recorder " + recorder.state);
recorder.requestData();
}
stream.oninactive = () => {
console.log("stream ended");
}
document.getElementById("start")
.onclick = () => {
if (recorder.state === "inactive") {
recorder.start();
} else {
recorder.resume();
}
console.log("recorder.state:", recorder.state);
}
document.getElementById("pause")
.onclick = () => {
if (recorder.state === "recording") {
recorder.pause();
}
console.log("recorder.state:", recorder.state);
}
document.getElementById("stop")
.onclick = () => {
if (recorder.state === "recording" || recorder.state === "paused") {
recorder.stop();
}
for (let track of stream.getTracks()) {
track.stop();
}
document.getElementById("start").onclick = null;
document.getElementById("pause").onclick = null;
console.log("recorder.state:", recorder.state
, "stream.active", stream.active);
}
})
.catch(err => {
console.error(err)
});
</script>
</body>
</html>

https://plnkr.co/edit/7caWYMsvub90G6pwDdQp?p=preview 普罗提

相关内容

最新更新

热门标签：