腾讯语音识别接口使用案例

1. 首先在腾讯ai开放平台注册,依次找到技术引擎->语音识别,选择语音识别一栏,在技术文档里左侧的新手指南下载python版本的SDK.同时在免费试用的接入能力中,创建应用,并得到app_key和app_id,修改test_aai_wxasrs.py

#-*- coding: UTF-8 -*-
import sys
import os
import optparse
import time
import apiutil
import base64
import json
import hashlib
import time
import glob
import codecs

app_key = ''#填入申请的app_key
app_id = '‘#填入申请的app_id

if __name__ == '__main__':
    wav_files = glob.glob(u'''./tools/TX_aiplat/audio/*.wav''')
    count = 1
    for wav_file in wav_files:
        if not os.path.exists(wav_file[:-3] + "trn"):
            seq = 0
            for_mat = 2
            rate = 16000
            bits = 16
            cont_res = 1                
            file_path = wav_file
            f = open(file_path, 'r')
            md5obj = hashlib.md5()  
            md5obj.update(f.read())  
            hash = md5obj.hexdigest()  
            speech_id = str(hash).upper() 
            f.close()
            f = open(file_path, 'rb')
            file_size = os.path.getsize(file_path)
            try:
                    chunk = f.read()
                    if not chunk:
                        pass
                        #break
                    else:
                        chunk_size = len(chunk)
                        if (seq + chunk_size) == file_size:
                            end = 1
                        else:
                            end = 0
                    ai_obj = apiutil.AiPlat(app_id, app_key)
                    print '----------------------SEND REQ----------------------'
                    rsp = ai_obj.getAaiWxAsrs(chunk, speech_id, end, for_mat, rate, bits, seq, chunk_size, cont_res)
                    seq += chunk_size
                    print rsp
                    if rsp['ret'] == 0:
                        results = json.dumps(rsp, encoding="UTF-8", ensure_ascii=False, sort_keys=False, indent=4)
                        results =  json.loads(results)
                        if results['data']['speech_text'] != u'':
                            print wav_file
                            trn_file = wav_file[:-3] + "trn"
                            with codecs.open(trn_file,'w','utf-8') as wp:
                                wp.write(results['data']['speech_text'])
                                count = count + 1
                            print u'''处理第%d个音频：%s''' %(count,wav_file)
                            print '----------------------API SUCC----------------------'
                            #time.sleep(1)
                        else:
                            os.remove(wav_file)
                    else:
                        print json.dumps(rsp, encoding="UTF-8", ensure_ascii=False, sort_keys=False, indent=4)
                        print '----------------------API FAIL----------------------'
            except:
                f.close()
                continue