清理了一些没有用到的vlm_detect文件

src/vlm_detect/vlm_backup/tts_node.py

@@ -0,0 +1,61 @@
+#!/usr/bin/env python3
+import rclpy, subprocess, os, wave
+from rclpy.node import Node
+from std_msgs.msg import String
+from piper import PiperVoice
+from piper.config import SynthesisConfig
+
+MODEL_PATH = '/home/sunrise/tts_model/zh_CN-huayan-medium.onnx'
+
+class TTSNode(Node):
+    def __init__(self):
+        super().__init__('tts_node')
+        self.declare_parameter('audio_sink', 'alsa_output.usb-C-Media_Electronics_Inc._USB_Audio_Device-00.analog-stereo')
+        self.declare_parameter('result_topic', '/vlm_result')
+        self.declare_parameter('tts_speed', 1.5)
+
+        self.audio_sink = self.get_parameter('audio_sink').value
+        result_topic = self.get_parameter('result_topic').value
+        tts_speed = self.get_parameter('tts_speed').value
+        self.length_scale = 1.0 / tts_speed
+        self.espeak_speed = int(175 * tts_speed)  # espeak default=175wpm, scale with tts_speed
+
+        self.sub = self.create_subscription(String, result_topic, self.callback, 10)
+        self.get_logger().info(f'Piper TTS started | voice=zh_CN-huayan | speed={tts_speed}x | sink={self.audio_sink}')
+        self.get_logger().info(f'Loading model: {MODEL_PATH}')
+        self.voice = PiperVoice.load(MODEL_PATH)
+        self.get_logger().info('Model loaded OK')
+
+    def callback(self, msg):
+        text = msg.data.strip()
+        if not text:
+            return
+        self.get_logger().info(f'TTS: {text}')
+        try:
+            syn_config = SynthesisConfig(length_scale=self.length_scale)
+            wav_path = '/tmp/tts_out.wav'
+            with wave.open(wav_path, 'wb') as wf:
+                self.voice.synthesize_wav(text, wf, syn_config=syn_config)
+            subprocess.Popen(
+                ['paplay', f'--device={self.audio_sink}', wav_path],
+                stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+        except Exception as e:
+            self.get_logger().error(f'Piper TTS failed, fallback espeak: {e}')
+            subprocess.Popen(
+                ['espeak-ng', '-v', 'cmn', '-s', str(self.espeak_speed), text],
+                env={**os.environ, 'PULSE_SINK': self.audio_sink},
+                stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+
+def main(args=None):
+    rclpy.init(args=args)
+    node = TTSNode()
+    try:
+        rclpy.spin(node)
+    except KeyboardInterrupt:
+        pass
+    finally:
+        node.destroy_node()
+        rclpy.shutdown()
+
+if __name__ == '__main__':
+    main()