腾讯云·社区
Jtro的技术分享:unity接入百度语音识别SDK
首先,你要去百度申请一个百度的AI开放平台的账号:https://ai.baidu.com/?track=cp:aipinzhuan|pf:pc|pp:AIpingtai|pu:title|ci:|kw:10005792 然后我们再申请一个android版的app.我们用到的只是这个的appid和appscred.
在创建的时候一定要选择安卓(敲黑板)
然后打开unity,新建脚本文件夹Scripts.然后再里面新建一个base脚本:
using System;
using System.Collections;
using UnityEngine;
using UnityEngine.Networking;
namespace Wit.BaiduAip.Speech
{
/// <summary>
/// 用户解析token的json数据
/// </summary>
[Serializable]
class TokenResponse
{
public string access_token = null;
}
public class Base
{
protected enum TokenFetchStatus
{
NotFetched,
Fetching,
Success,
Failed
}
public string SecretKey { get; private set; }
public string APIKey { get; private set; }
public string Token { get; private set; }
protected TokenFetchStatus tokenFetchStatus = TokenFetchStatus.NotFetched;
public Base(string apiKey, string secretKey)
{
APIKey = apiKey;
SecretKey = secretKey;
}
public IEnumerator GetAccessToken()
{
Debug.Log("[WitBaiduAip]Start fetching token...");
tokenFetchStatus = TokenFetchStatus.Fetching;
var uri =
string.Format(
"https://openapi.baidu.com/oauth/2.0/token?grant_type=client_credentials&client_id={0}&client_secret={1}",
APIKey, SecretKey);
var www = UnityWebRequest.Get(uri);
yield return www.SendWebRequest();
if (string.IsNullOrEmpty(www.error))
{
Debug.Log("[WitBaiduAip]" + www.downloadHandler.text);
var result = JsonUtility.FromJson<TokenResponse>(www.downloadHandler.text);
Token = result.access_token;
Debug.Log("[WitBaiduAip]Token has been fetched successfully");
tokenFetchStatus = TokenFetchStatus.Success;
}
else
{
Debug.LogError("[WitBaiduAip]"+www.error);
Debug.LogError("[WitBaiduAip]Token was fetched failed. Please check your APIKey and SecretKey");
tokenFetchStatus = TokenFetchStatus.Failed;
}
}
protected IEnumerator PreAction()
{
if (tokenFetchStatus == TokenFetchStatus.NotFetched)
{
Debug.Log("[WitBaiduAip]Token has not been fetched, now fetching...");
yield return GetAccessToken();
}
if (tokenFetchStatus == TokenFetchStatus.Fetching)
{
Debug.Log("[WitBaiduAip]Token is still being fetched, waiting...");
}
while (tokenFetchStatus == TokenFetchStatus.Fetching)
{
yield return null;
}
}
}
}这个是语音识别和语音听写转文字的父类.百度中语音听写和语音合成是免费的 新建一个Asr脚本
using System;
using System.Collections;
using UnityEngine;
using UnityEngine.Networking;
namespace Wit.BaiduAip.Speech
{
[Serializable]
public class AsrResponse
{
public int err_no;
public string err_msg;
public string sn;
public string[] result;
}
public class Asr : Base
{
private const string UrlAsr = @"https://vop.baidu.com/server_api";
public Asr(string apiKey, string secretKey) : base(apiKey, secretKey)
{
}
public IEnumerator Recognize(byte[] data, Action<AsrResponse> callback)
{
yield return PreAction ();
if (tokenFetchStatus == Base.TokenFetchStatus.Failed) {
Debug.LogError("Token fetched failed, please check your APIKey and SecretKey");
yield break;
}
var uri = string.Format("{0}?lan=zh&cuid={1}&token={2}", UrlAsr, SystemInfo.deviceUniqueIdentifier, Token);
var form = new WWWForm();
form.AddBinaryData("audio", data);
var www = UnityWebRequest.Post(uri, form);
www.SetRequestHeader("Content-Type", "audio/pcm;rate=16000");
yield return www.SendWebRequest();
if (string.IsNullOrEmpty(www.error))
{
Debug.Log("[WitBaiduAip]"+www.downloadHandler.text);
callback(JsonUtility.FromJson<AsrResponse>(www.downloadHandler.text));
}
else
Debug.LogError(www.error);
}
/// <summary>
/// 将Unity的AudioClip数据转化为PCM格式16bit数据
/// </summary>
/// <param name="clip"></param>
/// <returns></returns>
public static byte[] ConvertAudioClipToPCM16(AudioClip clip)
{
var samples = new float[clip.samples * clip.channels];
clip.GetData(samples, 0);
var samples_int16 = new short[samples.Length];
for (var index = 0; index < samples.Length; index++)
{
var f = samples[index];
samples_int16[index] = (short) (f * short.MaxValue);
}
var byteArray = new byte[samples_int16.Length * 2];
Buffer.BlockCopy(samples_int16, 0, byteArray, 0, byteArray.Length);
return byteArray;
}
}
}再新建一个Tts
using System;
using System.Collections;
using System.Collections.Generic;
using System.Text;
using UnityEngine;
using UnityEngine.Networking;
namespace Wit.BaiduAip.Speech
{
/// <summary>
/// 语音合成结果
/// </summary>
[Serializable]
public class TtsResponse
{
public int err_no;
public string err_msg;
public string sn;
public int idx;
public bool Success
{
get { return err_no == 0; }
}
public AudioClip clip;
}
public class Tts : Base
{
public enum Pronouncer
{
Female, // 0为普通女声
Male, // 1为普通男生
Duxiaoyao, // 3为情感合成-度逍遥
Duyaya // 4为情感合成-度丫丫
}
private const string UrlTts = "http://tsn.baidu.com/text2audio";
public Tts(string apiKey, string secretKey) : base(apiKey, secretKey)
{
}
public IEnumerator Synthesis(string text, Action<TtsResponse> callback, int speed = 5, int pit = 5, int vol = 5,
Pronouncer per = Pronouncer.Female)
{
yield return PreAction();
if (tokenFetchStatus == Base.TokenFetchStatus.Failed)
{
Debug.LogError("Token was fetched failed. Please check your APIKey and SecretKey");
callback(new TtsResponse()
{
err_no = -1,
err_msg = "Token was fetched failed. Please check your APIKey and SecretKey"
});
yield break;
}
var param = new Dictionary<string, string>();
param.Add("tex", text);
param.Add("tok", Token);
param.Add("cuid", SystemInfo.deviceUniqueIdentifier);
param.Add("ctp", "1");
param.Add("lan", "zh");
param.Add("spd", Mathf.Clamp(speed, 0, 9).ToString());
param.Add("pit", Mathf.Clamp(pit, 0, 9).ToString());
param.Add("vol", Mathf.Clamp(vol, 0, 15).ToString());
param.Add("per", ((int)per).ToString());
#if UNITY_STANDALONE || UNITY_EDITOR || UNITY_UWP
param.Add("aue", "6"); // set to wav, default is mp3
#endif
string url = UrlTts;
int i = 0;
foreach (var p in param)
{
url += i != 0 ? "&" : "?";
url += p.Key + "=" + p.Value;
i++;
}
#if UNITY_STANDALONE || UNITY_EDITOR || UNITY_UWP
var www = UnityWebRequestMultimedia.GetAudioClip(url, AudioType.WAV);
#else
var www = UnityWebRequestMultimedia.GetAudioClip(url, AudioType.MPEG);
#endif
Debug.Log("[WitBaiduAip]" + www.url);
yield return www.SendWebRequest();
if (string.IsNullOrEmpty(www.error))
{
var type = www.GetResponseHeader("Content-Type");
Debug.Log("[WitBaiduAip]response type: " + type);
if (type.Contains("audio"))
{
#if UNITY_STANDALONE || UNITY_EDITOR || UNITY_UWP
var clip = DownloadHandlerAudioClip.GetContent(www);
var response = new TtsResponse { clip = clip };
#else
var response = new TtsResponse {clip = DownloadHandlerAudioClip.GetContent(www) };
#endif
callback(response);
}
else
{
var textBytes = www.downloadHandler.data;
var errorText = Encoding.UTF8.GetString(textBytes);
Debug.LogError("[WitBaiduAip]" + errorText);
callback(JsonUtility.FromJson<TtsResponse>(errorText));
}
}
else
{
Debug.LogError(www.error);
}
}
}
}到这里我们的前期工作就做完了,接下来就是再场景中使用了 首先新建一个场景语音识别,场景中2个按钮一个text文本
using UnityEngine;
using UnityEngine.UI;
using Wit.BaiduAip.Speech;
public class AsrDemo : MonoBehaviour
{
public string APIKey = "";
public string SecretKey = "";
public Button StartButton;
public Button StopButton;
public Text DescriptionText;
private AudioClip _clipRecord;
private Asr _asr;
// Microphone is not supported in Webgl
#if !UNITY_WEBGL
void Start()
{
_asr = new Asr(APIKey, SecretKey);
StartCoroutine(_asr.GetAccessToken());
StartButton.gameObject.SetActive(true);
StopButton.gameObject.SetActive(false);
DescriptionText.text = "";
StartButton.onClick.AddListener(OnClickStartButton);
StopButton.onClick.AddListener(OnClickStopButton);
}
private void OnClickStartButton()
{
StartButton.gameObject.SetActive(false);
StopButton.gameObject.SetActive(true);
DescriptionText.text = "Listening...";
_clipRecord = Microphone.Start(null, false, 30, 16000);
}
private void OnClickStopButton()
{
StartButton.gameObject.SetActive(false);
StopButton.gameObject.SetActive(false);
DescriptionText.text = "Recognizing...";
Microphone.End(null);
Debug.Log("[WitBaiduAip demo]end record");
var data = Asr.ConvertAudioClipToPCM16(_clipRecord);
StartCoroutine(_asr.Recognize(data, s =>
{
DescriptionText.text = s.result != null && s.result.Length > 0 ? s.result[0] : "未识别到声音";
StartButton.gameObject.SetActive(true);
}));
}
#endif
}然后拖拽到摄像机上,配置脚本的引用文件
然后现在你可以测试你的场景了,可以识别到你说的话. 下面是语音合成,新建一个场景然后新键脚本文件
using UnityEngine;
using UnityEngine.UI;
using Wit.BaiduAip.Speech;
public class TtsDemo : MonoBehaviour
{
public string APIKey = "";
public string SecretKey = "";
public Button SynthesisButton;
public InputField Input;
public Text DescriptionText;
private Tts _asr;
private AudioSource _audioSource;
private bool _startPlaying;
void Start()
{
_asr = new Tts(APIKey, SecretKey);
StartCoroutine(_asr.GetAccessToken());
_audioSource = gameObject.AddComponent<AudioSource>();
DescriptionText.text = "";
SynthesisButton.onClick.AddListener(OnClickSynthesisButton);
}
private void OnClickSynthesisButton()
{
SynthesisButton.gameObject.SetActive(false);
DescriptionText.text = "合成中...";
StartCoroutine(_asr.Synthesis(Input.text, s =>
{
if (s.Success)
{
DescriptionText.text = "合成成功，正在播放";
_audioSource.clip = s.clip;
_audioSource.Play();
_startPlaying = true;
}
else
{
DescriptionText.text = s.err_msg;
SynthesisButton.gameObject.SetActive(true);
}
}));
}
void Update()
{
if (_startPlaying)
{
if (!_audioSource.isPlaying)
{
_startPlaying = false;
DescriptionText.text = "播放完毕，可以修改文本继续测试";
SynthesisButton.gameObject.SetActive(true);
}
}
}
}场景中加入:输入框和一个按钮 把脚本拖拽到摄像机上 脚本配置如下:
然后运行的时候在输入框中输入你要合成的文字,点击按钮之后可以听到合成的声音.
本文参与腾讯云自媒体分享计划，欢迎正在阅读的你也加入，一起分享。
相关文章
扫码关注云+社区
领取腾讯云代金券
我来说两句