modify param

dataset/audio/segments/test1_segment_1.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:877aee33d778b34af2f0b819ac822d80316e97b73cb3823c1f436dbef8efcb0e
-size 35564
+oid sha256:28c322856596e185c7405464090742ad114d1c2830bbc2b86d410c252534bb14
+size 21164

dataset/audio/segments/test1_segment_10.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3135d983a5260d846e6cf165583efa3a0ef379bd86c885e678a63b41f66f548b
-size 48044
+oid sha256:b29c94ca02c45ee8145dade0866f0f1c2451af7a11ee3b1b98b11fb867ec1a2a
+size 97324

dataset/audio/segments/test1_segment_11.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a966cbb2e82ebd278692adad509a18061306b73b715fc4a93468c27ed61627b
-size 111404
+oid sha256:3c923e80fb6189b28fdcc6d901fe53c7e5af4410fd41bbe2ab49d1f7386c3df6
+size 39084

dataset/audio/segments/test1_segment_12.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:52cfbcdc17cc5f190df467310f1a91c89e27f79662b2ce13f4ff5ec07015afec
-size 71084
+oid sha256:82984e994f0a59fe2cbaa02ec6805946b8b982e92d6d785395e472cc792eb5c8
+size 95404

dataset/audio/segments/test1_segment_13.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cbc350f2224afe3c79ea985efd38ec02ae67c7b7c645327a51004f82ecabd1f
+size 35884

dataset/audio/segments/test1_segment_14.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc8b91e09a1097f8477fe5180891d6fdbf9a150e52b983cf600fd33aa0adc298
+size 32044

dataset/audio/segments/test1_segment_15.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:840350c64d3056851547ecd8a81952b87b064cc3d87e037f084aa9ce77d27c89
+size 49324

dataset/audio/segments/test1_segment_16.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d124dd2d7d788dfb01168267f124baa9d176a8656561dfc080aa402130055f0
+size 113964

dataset/audio/segments/test1_segment_17.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a9e2067b592e25358eb1a35728fb9319fd0472230f1184c1d11af6195e050d2
+size 73644

dataset/audio/segments/test1_segment_2.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:81375721eb3a532941083c9781f53f5e0f1ccbe1ef4108f98a019de400f5c564
-size 117164
+oid sha256:8618632e403ddef0f05b160710ed6eab17a5d4b7042b5cdda435212c57e5efcf
+size 39084

dataset/audio/segments/test1_segment_3.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bd6120ff04e7365640b9e3a1fb062bc1c31ce0dc54904bd27e25ac5a0b068cde
-size 149804
+oid sha256:63859aead158b66a043a6fd16496a49211ae3516378c894174adfccb04a218b7
+size 22444

dataset/audio/segments/test1_segment_4.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:99bc0d18ffd0d10742b8d6b5450e537eccd1497c2247e714fa8efe6beb602abd
-size 41324
+oid sha256:04e1bb445589f502605ec58654c08f2e66958973ff49a9f408933bffb16b9ee0
+size 116524

dataset/audio/segments/test1_segment_5.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a9e2196db3537028898b87442f074523251b33219302e6eb8518fb33396c30bd
-size 122924
+oid sha256:1acd47e3c4c68dcf959cf62d9faf988507e90f21145ca33a1d8829314d2d01a9
+size 126124

dataset/audio/segments/test1_segment_6.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e929f7966a425a559b7442a2914cb99b0df74f1d02938264642dc71f160fc383
-size 113324
+oid sha256:b94d1a2e269cf95774b15fd75f43fc72ccd1a1d2e57985be0bb20334603532d7
+size 46124

dataset/audio/segments/test1_segment_7.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:65d800356647c415d80e59fac63db01df31ce51a497aacf43f98aa0e6ec468cb
-size 77804
+oid sha256:b075f47a9499e873430eaf99f5e0149e0f52e3fe7e1172be09fa4ce9eb56d7ae
+size 42924

dataset/audio/segments/test1_segment_8.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c1c574a7c20332f85c6260febf6eae232473a798404ca29f1b54ac39e5b2d35c
-size 91244
+oid sha256:4eb3a808a08cdac833973dec3811d2bfac7258b6259dfaf5f7a0c69ed6cef898
+size 129964

dataset/audio/segments/test1_segment_9.wav

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f943b20eb3aafa0befb884f5d125e0596d3f419d8a3c5546ff3cf878603c36b8
-size 67244
+oid sha256:15079fb8b683f5d8b3cad064bcd21488c3acf9837a2dc7095c922df3dd2a3f4b
+size 61484

dataset/transcripts/test1_segment_1_20250423_162821.json

@@ -0,0 +1,177 @@
+{
+  "audio_file": "dataset/audio/segments\\test1_segment_1.wav",
+  "timestamp": "20250423_162821",
+  "segments": [
+    {
+      "text": "第一單元",
+      "start_time": 3.26,
+      "end_time": 3.94,
+      "confidence": 0.661865234375,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "音频数据处理",
+      "start_time": 4.34,
+      "end_time": 6.34,
+      "confidence": 0.36962890625,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "单元简介",
+      "start_time": 7.1,
+      "end_time": 7.859999999999999,
+      "confidence": 0.77685546875,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "所有音频或语音相关的任务都需要使用音频文件",
+      "start_time": 8.8,
+      "end_time": 12.4,
+      "confidence": 0.9898605346679688,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "在我们先入了解这些任务之前",
+      "start_time": 12.8,
+      "end_time": 14.8,
+      "confidence": 0.83056640625,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "我们需要了解音频文件的实际内容",
+      "start_time": 14.8,
+      "end_time": 16.8,
+      "confidence": 0.83056640625,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "以及如何利用一篇文件",
+      "start_time": 17.32,
+      "end_time": 18.72,
+      "confidence": 0.9420166015625,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "本来员将会你介绍的",
+      "start_time": 19.76,
+      "end_time": 21.040000000000003,
+      "confidence": 0.90777587890625,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "本单员将为你介绍于音频数据相关的基本概念",
+      "start_time": 21.62,
+      "end_time": 25.62,
+      "confidence": 0.92706298828125,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "包括波形、彩虹、綠和平不圖",
+      "start_time": 26.28,
+      "end_time": 28.28,
+      "confidence": 0.944183349609375,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "你会学习到如何使用音频输语集",
+      "start_time": 28.56,
+      "end_time": 30.36,
+      "confidence": 0.98211669921875,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "包括音频输语夹仔",
+      "start_time": 30.36,
+      "end_time": 31.64,
+      "confidence": 0.98211669921875,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "音频书记预处理一期",
+      "start_time": 31.98,
+      "end_time": 33.26,
+      "confidence": 0.9024658203125,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "高效加载大规模音频数级的流适加载方法",
+      "start_time": 33.54,
+      "end_time": 36.54,
+      "confidence": 0.94915771484375,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "完成本單元的學期後",
+      "start_time": 37.82,
+      "end_time": 38.98,
+      "confidence": 0.9828643798828125,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "你会掌握",
+      "start_time": 39.34,
+      "end_time": 40.34,
+      "confidence": 0.8743896484375,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "基础的音频相关数",
+      "start_time": 40.86,
+      "end_time": 42.46,
+      "confidence": 0.75927734375,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "并且掌握针对不同应用的音频数据处理工具",
+      "start_time": 43.04,
+      "end_time": 46.64,
+      "confidence": 0.9911956787109375,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    },
+    {
+      "text": "本难援的知识会成为后面章节的基础",
+      "start_time": 47.5,
+      "end_time": 49.94,
+      "confidence": 0.954833984375,
+      "verified": false,
+      "verified_text": null,
+      "verification_notes": null
+    }
+  ]
+}

vad/audio_processor.py

@@ -1,3 +1,4 @@
+import sys
 import numpy as np
 import soundfile as sf
 from typing import List, Tuple, Optional, Dict
@@ -18,11 +19,11 @@ class AudioSegment:
 class AudioProcessor:
     def __init__(self, 
                  sample_rate: int = 16000,
-                 frame_duration_ms: int = 30,
-                 vad_level: int = 1,          # 降低VAD灵敏度
-                 min_silence_duration: float = 0.5,  # 静音持续时间
-                 min_speech_duration: float = 1.0,   # 增加最小语音持续时间，确保完整句子
-                 amplitude_threshold: float = 0.003): # 进一步降低振幅阈值
+                 frame_duration_ms: int = 20,
+                 vad_level: int = 0,          # 降低VAD灵敏度
+                 min_silence_duration: float = 0.3,  # 静音持续时间
+                 min_speech_duration: float = 0.3,   # 增加最小语音持续时间，确保完整句子
+                 amplitude_threshold: float = 0.0015): # 进一步降低振幅阈值
         """
         初始化音频处理器
         
@@ -160,6 +161,9 @@ class AudioProcessor:
         print(f"总帧数: {total_frames}")
         print(f"语音帧数: {speech_frames}")
         print(f"检测到的语音片段数: {len(segments)}")
+        # print("中断程序以进行调试")
+        # sys.exit(1)
+
         
         # 保存中间结果到临时文件
         temp_dir = "../dataset/audio/temp"