File size: 8,815 Bytes
27e74f3 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 | import webrtcvad
import numpy as np
import threading
import time
import logging
import pyaudio
from src.constants.constants import AbortReason, DeviceState
# 配置日志
logger = logging.getLogger("VADDetector")
class VADDetector:
"""基于WebRTC VAD的语音活动检测器，用于检测用户打断"""
def __init__(self, audio_codec, protocol, app_instance, loop):
"""初始化VAD检测器
参数:
audio_codec: 音频编解码器实例
protocol: 通信协议实例
app_instance: 应用程序实例
loop: 事件循环
"""
self.audio_codec = audio_codec
self.protocol = protocol
self.app = app_instance
self.loop = loop
# VAD设置
self.vad = webrtcvad.Vad()
self.vad.set_mode(3) # 设置最高灵敏度
# 参数设置
self.sample_rate = 16000
self.frame_duration = 20 # 毫秒
self.frame_size = int(self.sample_rate * self.frame_duration / 1000)
self.speech_window = 5 # 连续检测到多少帧语音才触发打断
self.energy_threshold = 300 # 能量阈值
# 状态变量
self.running = False
self.paused = False
self.thread = None
self.speech_count = 0
self.silence_count = 0
self.triggered = False
# 创建独立的PyAudio实例和流，避免与主音频流冲突
self.pa = None
self.stream = None
def start(self):
"""启动VAD检测器"""
if self.thread and self.thread.is_alive():
logger.warning("VAD检测器已经在运行")
return
self.running = True
self.paused = False
# 初始化PyAudio和流
self._initialize_audio_stream()
# 启动检测线程
self.thread = threading.Thread(target=self._detection_loop, daemon=True)
self.thread.start()
logger.info("VAD检测器已启动")
def stop(self):
"""停止VAD检测器"""
self.running = False
# 关闭音频流
self._close_audio_stream()
if self.thread and self.thread.is_alive():
self.thread.join(timeout=1.0)
logger.info("VAD检测器已停止")
def pause(self):
"""暂停VAD检测"""
self.paused = True
logger.info("VAD检测器已暂停")
def resume(self):
"""恢复VAD检测"""
self.paused = False
# 重置状态
self.speech_count = 0
self.silence_count = 0
self.triggered = False
logger.info("VAD检测器已恢复")
def is_running(self):
"""检查VAD检测器是否正在运行"""
return self.running and not self.paused
def _initialize_audio_stream(self):
"""初始化独立的音频流"""
try:
# 创建PyAudio实例
self.pa = pyaudio.PyAudio()
# 获取默认输入设备
device_index = None
for i in range(self.pa.get_device_count()):
device_info = self.pa.get_device_info_by_index(i)
if device_info['maxInputChannels'] > 0:
device_index = i
break
if device_index is None:
logger.error("找不到可用的输入设备")
return False
# 创建输入流
self.stream = self.pa.open(
format=pyaudio.paInt16,
channels=1,
rate=self.sample_rate,
input=True,
input_device_index=device_index,
frames_per_buffer=self.frame_size,
start=True
)
logger.info(f"VAD检测器音频流已初始化，使用设备索引: {device_index}")
return True
except Exception as e:
logger.error(f"初始化VAD音频流失败: {e}")
return False
def _close_audio_stream(self):
"""关闭音频流"""
try:
if self.stream:
self.stream.stop_stream()
self.stream.close()
self.stream = None
if self.pa:
self.pa.terminate()
self.pa = None
logger.info("VAD检测器音频流已关闭")
except Exception as e:
logger.error(f"关闭VAD音频流失败: {e}")
def _detection_loop(self):
"""VAD检测主循环"""
logger.info("VAD检测循环已启动")
while self.running:
# 如果暂停或者音频流未初始化，则跳过
if self.paused or not self.stream:
time.sleep(0.1)
continue
try:
# 只在说话状态下进行检测
if self.app.device_state == DeviceState.SPEAKING:
# 读取音频帧
frame = self._read_audio_frame()
if not frame:
time.sleep(0.01)
continue
# 检测是否是语音
is_speech = self._detect_speech(frame)
# 如果检测到语音并且达到触发条件，处理打断
if is_speech:
self._handle_speech_frame(frame)
else:
self._handle_silence_frame(frame)
else:
# 不在说话状态，重置状态
self._reset_state()
except Exception as e:
logger.error(f"VAD检测循环出错: {e}")
time.sleep(0.01) # 小延迟，减少CPU使用
logger.info("VAD检测循环已结束")
def _read_audio_frame(self):
"""读取一帧音频数据"""
try:
if not self.stream or not self.stream.is_active():
return None
# 读取音频数据
data = self.stream.read(self.frame_size, exception_on_overflow=False)
return data
except Exception as e:
logger.error(f"读取音频帧失败: {e}")
return None
def _detect_speech(self, frame):
"""检测是否是语音"""
try:
# 确保帧长度正确
if len(frame) != self.frame_size * 2: # 16位音频，每个样本2字节
return False
# 使用VAD检测
is_speech = self.vad.is_speech(frame, self.sample_rate)
# 计算音频能量
audio_data = np.frombuffer(frame, dtype=np.int16)
energy = np.mean(np.abs(audio_data))
# 结合VAD和能量阈值
is_valid_speech = is_speech and energy > self.energy_threshold
if is_valid_speech:
logger.debug(f'检测到语音 [能量: {energy:.2f}] [连续语音帧: {self.speech_count+1}]')
return is_valid_speech
except Exception as e:
logger.error(f"检测语音失败: {e}")
return False
def _handle_speech_frame(self, frame):
"""处理语音帧"""
self.speech_count += 1
self.silence_count = 0
# 检测到足够的连续语音帧，触发打断
if self.speech_count >= self.speech_window and not self.triggered:
self.triggered = True
logger.info("检测到持续语音，触发打断！")
self._trigger_interrupt()
# 立即暂停自己，防止重复触发
self.paused = True
logger.info("VAD检测器已自动暂停以防止重复触发")
# 重置状态
self.speech_count = 0
self.silence_count = 0
self.triggered = False
def _handle_silence_frame(self, frame):
"""处理静音帧"""
self.silence_count += 1
self.speech_count = 0
def _reset_state(self):
"""重置状态"""
self.speech_count = 0
self.silence_count = 0
self.triggered = False
def _trigger_interrupt(self):
"""触发打断"""
# 通知应用程序中止当前语音输出
self.app.schedule(lambda: self.app.abort_speaking(AbortReason.WAKE_WORD_DETECTED))
|