语音识别bug

#       如果消息为200,则表示talker在交互过程中未识别到用户的语音输入,开启人脸验证确认用户是否还在面前(排除噪声的影响)
#       确认时间为3秒,3秒都没有人脸,则确认用户不在.确认用户还在,则提醒对方重新说一下,如果不在,则取消交互,进入等待人脸介入环节.
#        TODO:有人的话确认是否为本轮交互的同一人.
        elif loop_count.data == :#unrecognized voice, reconfig if there still people talking to
            start_time = time.time()
            while True:
                # Capture frame-by-frame
                ret, frame = cap.read()
                if ret:
                    cv2.imwrite('tmp.jpg',frame)
                with open("tmp.jpg", "rb") as fp:
                    image_binary = fp.read()
                    image_binary = base64.b64encode(image_binary)
                    post_data = {"Image":image_binary}
                    body = JSONEncoder().encode(post_data)
                    req = urllib2.Request(url, body)
                    response = urllib2.urlopen(req).read()
                    body = JSONDecoder().decode(response)
                    rospy.loginfo(body)

                if body['Id'] == 'UNKNOWN' or body['Id'] == 'None':#reconfiged nobody
                    msg.is_staff =
                    msg.name = 'nobody'
                    #self.face_result_pub.publish(msg)
                    print "nobody"
                    now_time = time.time()
                    if now_time - start_time <= :
                        continue
                    else:
                            self.face_result_pub.publish(msg)
                        break
                else:#reconfiged unregistered people
                    msg.is_staff =
                    msg.name = 'unknown'
                    self.face_result_pub.publish(msg)
                    break
        else:
            pass

修改直到3秒结束，再发布 self.face_result_pub.publish(msg)