iOS語音聽寫、語音合成
阿新 • • 發佈：2019-01-22
本文語音聽寫以及語音合成基於訊飛科技提供的sdk
根據官方文件，下載sdk，建立自己的應用，獲取appid。
sdk放進自己的工程裡面，然後新增需要的庫。
編譯一下 ，如果編譯報錯，修改Targets -》 Build Settings -》Build Options -》Enable Bitcode 設定為No 編譯通過，sdk環境整合成功。
在程式入口，配置使用者資訊：
- (BOOL)application:(UIApplication *)application didFinishLaunchingWithOptions:(NSDictionary *)launchOptions { NSString *initString = [[NSString alloc] initWithFormat:@"appid=%@",@"000000"];//此處填寫自己申請應用的appId [IFlySpeechUtility createUtility:initString]; }
一、語音合成，也就是閱讀文字，把文字訊息轉化成語音輸出：
1.匯入相關的標頭檔案
#import <iflyMSC/IFlySpeechSynthesizerDelegate.h>
#import <iflyMSC/IFlySpeechSynthesizer.h>
#import <iflyMSC/IFlySpeechConstant.h>@interface ViewController ()<IFlySpeechSynthesizerDelegate> @property (strong, nonatomic) IFlySpeechSynthesizer *synthesizer; @property (nonatomic , strong) UITextView *textView; @property (nonatomic , strong) UIButton *startBtn; @property (nonatomic , strong) UIButton *cancelBtn; @end
3.建立文字識別物件，
self.synthesizer = [IFlySpeechSynthesizer sharedInstance];
self.synthesizer.delegate = self;並可以自定義設定該物件的關鍵屬性，不設定將會是預設值
[self.synthesizer setParameter:@"50" forKey:[IFlySpeechConstant SPEED]]; [self.synthesizer setParameter:@"50" forKey:[IFlySpeechConstant VOLUME]]; [self.synthesizer setParameter:@"VIXR" forKey:[IFlySpeechConstant VOICE_NAME]]; [self.synthesizer setParameter:@"8000" forKey:[IFlySpeechConstant SAMPLE_RATE]]; [self.synthesizer setParameter:@"temp.pcm" forKey:[IFlySpeechConstant TTS_AUDIO_PATH]]; [self.synthesizer setParameter:@"custom" forKey:[IFlySpeechConstant PARAMS]];
4.功能啟動的呼叫方法
- (void)startBtnCliked {
[self.synthesizer startSpeaking:self.textView.text];
}- (void)onCompleted:(IFlySpeechError *)error {
NSLog(@"講完了");
}二、語音識別，將語音翻譯成文字訊息。
sdk中提供了兩種方式，一種是帶介面的語音識別，也就是有一個小話筒的介面效果。另一種是無介面的。
先說有介面的，也比較簡單。
1.匯入標頭檔案
#import "iflyMSC/IFlyMSC.h"
#import "ISRDataHelper.h"2.遵守協議
@interface ViewController ()<IFlyRecognizerViewDelegate> {
IFlyRecognizerView *_iflyRecognizerView;
}
@property (nonatomic , strong) UITextView *textView;
@property (nonatomic , strong) UIButton *startBtn;
@property (nonatomic , strong) UIButton *cancelBtn;
@end3.建立物件
_iflyRecognizerView = [[IFlyRecognizerView alloc] initWithCenter:self.view.center];
_iflyRecognizerView.delegate = self;
[_iflyRecognizerView setParameter: @"iat" forKey: [IFlySpeechConstant IFLY_DOMAIN]];
[_iflyRecognizerView setParameter:@"asrview.pcm " forKey:[IFlySpeechConstant ASR_AUDIO_PATH]];4.開始語音識別
- (void)startBtnCliked {
[_textView resignFirstResponder];
//啟動識別服務
[_iflyRecognizerView start];
}5.實現代理回撥，處理識別的結果
/*識別結果返回代理
@param resultArray 識別結果
@ param isLast 表示是否最後一次結果
*/
- (void)onResult: (NSArray *)resultArray isLast:(BOOL) isLast
{
NSMutableString *resultString = [[NSMutableString alloc] init];
NSDictionary *dic = resultArray[0];
for (NSString *key in dic) {
[resultString appendFormat:@"%@",key];
}
NSString * resultFromJson = [ISRDataHelper stringFromJson:resultString];
_textView.text = [NSString stringWithFormat:@"%@%@", _textView.text,resultFromJson];
}
/*識別會話錯誤返回代理
@ param error 錯誤碼
*/
- (void)onError: (IFlySpeechError *) error
{
NSLog(@"%s",__func__);
NSString *text ;
if (error.errorCode ==0 ) {
}
else
{
text = [NSString stringWithFormat:@"發生錯誤：%d %@",error.errorCode,error.errorDesc];
NSLog(@"%@",text);
[_iflyRecognizerView cancel];
}
self.startBtn.enabled = YES;
}
然後是不帶介面的語音識別。
1.同樣匯入標頭檔案
#import <iflyMSC/IFlySpeechRecognizerDelegate.h>
#import <iflyMSC/IFlySpeechRecognizer.h>
#import <iflyMSC/IFlySpeechError.h>
#import "ISRDataHelper.h"2.建立物件 遵守協議
@interface ViewController ()<IFlySpeechRecognizerDelegate>
@property (nonatomic , strong) IFlySpeechRecognizer *iFlySpeechRecognizer;
@property (nonatomic , strong) UITextView *textView;
@property (nonatomic , strong) UIButton *startBtn;
@property (nonatomic , strong) UIButton *cancelBtn;
@end self.iFlySpeechRecognizer = [IFlySpeechRecognizer sharedInstance];
self.iFlySpeechRecognizer.delegate = self;3.按鈕方法，功能啟動入口
- (void)startBtnCliked {
BOOL ret = [self.iFlySpeechRecognizer startListening];
if (ret) {
NSLog(@"成功");
} else {
NSLog(@"啟動失敗");
}
}
- (void)cancelBtnCliked {
[self.iFlySpeechRecognizer cancel];
}4.實現代理回撥方法
- (void) onError:(IFlySpeechError *)errorCode {
NSLog(@"%@",errorCode.errorDesc);
}
- (void) onResults:(NSArray *) results isLast:(BOOL)isLast {
NSMutableString *resultString = [[NSMutableString alloc] init];
NSDictionary *dic = results[0];
for (NSString *key in dic) {
[resultString appendFormat:@"%@",key];
}
NSString * resultFromJson = [ISRDataHelper stringFromJson:resultString];
_textView.text = [NSString stringWithFormat:@"%@%@", _textView.text,resultFromJson];
}這個ISRDataHelper是訊飛demo裡面封裝好的一個類，用來處理翻譯過來的文字。實際上可以這樣解析預設的json串：
- (void) onResults:(NSArray *) results isLast:(BOOL)isLast {
NSArray * temp = [[NSArray alloc]init];
NSString * str = [[NSString alloc]init];
NSMutableString *result = [[NSMutableString alloc] init];
NSDictionary *dic = results[0];
for (NSString *key in dic) {
[result appendFormat:@"%@",key];
}
NSLog(@"聽寫結果：%@",result);
//---------訊飛語音識別JSON資料解析---------//
NSError * error;
NSData * data = [result dataUsingEncoding:NSUTF8StringEncoding];
NSLog(@"data: %@",data);
NSDictionary * dic_result =[NSJSONSerialization JSONObjectWithData:data options:NSJSONReadingMutableLeaves error:&error];
NSArray * array_ws = [dic_result objectForKey:@"ws"];
//遍歷識別結果的每一個單詞
for (int i=0; i<array_ws.count; i++) {
temp = [[array_ws objectAtIndex:i] objectForKey:@"cw"];
NSDictionary * dic_cw = [temp objectAtIndex:0];
str = [str stringByAppendingString:[dic_cw objectForKey:@"w"]];
NSLog(@"識別結果:%@",[dic_cw objectForKey:@"w"]);
}
NSLog(@"最終的識別結果:%@",str);
//去掉識別結果最後的標點符號
if ([str isEqualToString:@"。"] || [str isEqualToString:@"？"] || [str isEqualToString:@"！"]) {
NSLog(@"末尾標點符號：%@",str);
}
else{
self.textView.text = str;
}
}
目前只整理了這兩個功能的程式碼，後續需要會再研究其他功能的程式碼。