初始可用版本

**已知问题**
- 无法正确区分说话人
- 语音识别精度有待提高

.gitignore

@@ -0,0 +1,56 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# 虚拟环境
+funasr_env/
+venv/
+env/
+ENV/
+
+# 模型缓存（体积较大）
+models/
+
+# 测试输出
+*_result.json
+*_result.srt
+*.log
+
+# IDE
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+
+# 系统文件
+.DS_Store
+Thumbs.db
+
+# 输入输出
+output/
+input/
+
+# 音频文件（可选，根据需要调整）
+# *.wav
+# *.mp3
+# *.m4a
+# *.flac

CLAUDE.md

@@ -0,0 +1,65 @@
+# CLAUDE.md
+
+Behavioral guidelines to reduce common LLM coding mistakes. Merge with project-specific instructions as needed.
+
+**Tradeoff:** These guidelines bias toward caution over speed. For trivial tasks, use judgment.
+
+## 1. Think Before Coding
+
+**Don't assume. Don't hide confusion. Surface tradeoffs.**
+
+Before implementing:
+- State your assumptions explicitly. If uncertain, ask.
+- If multiple interpretations exist, present them - don't pick silently.
+- If a simpler approach exists, say so. Push back when warranted.
+- If something is unclear, stop. Name what's confusing. Ask.
+
+## 2. Simplicity First
+
+**Minimum code that solves the problem. Nothing speculative.**
+
+- No features beyond what was asked.
+- No abstractions for single-use code.
+- No "flexibility" or "configurability" that wasn't requested.
+- No error handling for impossible scenarios.
+- If you write 200 lines and it could be 50, rewrite it.
+
+Ask yourself: "Would a senior engineer say this is overcomplicated?" If yes, simplify.
+
+## 3. Surgical Changes
+
+**Touch only what you must. Clean up only your own mess.**
+
+When editing existing code:
+- Don't "improve" adjacent code, comments, or formatting.
+- Don't refactor things that aren't broken.
+- Match existing style, even if you'd do it differently.
+- If you notice unrelated dead code, mention it - don't delete it.
+
+When your changes create orphans:
+- Remove imports/variables/functions that YOUR changes made unused.
+- Don't remove pre-existing dead code unless asked.
+
+The test: Every changed line should trace directly to the user's request.
+
+## 4. Goal-Driven Execution
+
+**Define success criteria. Loop until verified.**
+
+Transform tasks into verifiable goals:
+- "Add validation" → "Write tests for invalid inputs, then make them pass"
+- "Fix the bug" → "Write a test that reproduces it, then make it pass"
+- "Refactor X" → "Ensure tests pass before and after"
+
+For multi-step tasks, state a brief plan:
+```
+1. [Step] → verify: [check]
+2. [Step] → verify: [check]
+3. [Step] → verify: [check]
+```
+
+Strong success criteria let you loop independently. Weak criteria ("make it work") require constant clarification.
+
+---
+
+**These guidelines are working if:** fewer unnecessary changes in diffs, fewer rewrites due to overcomplication, and clarifying questions come before implementation rather than after mistakes.

README.md

@@ -0,0 +1,230 @@
+# FunASR 语音识别服务
+
+基于阿里达摩院 [FunASR](https://github.com/alibaba-damo-academy/FunASR) 的本地语音识别解决方案。
+
+## 功能特性
+
+| 功能 | 说明 |
+|------|------|
+| **句级时间戳** | 每句话的开始和结束时间 |
+| **说话人分离** | 自动区分不同说话人 |
+| **抗噪能力** | VAD 语音活动检测，过滤噪音 |
+| **本地部署** | 完全离线运行，数据不上传云端 |
+
+## 项目结构
+
+```
+audio2/
+├── funasr_env/              # 虚拟环境
+├── models/                  # 模型缓存目录
+├── asr_service.py           # 核心服务类
+├── test_asr.py              # 测试脚本
+├── example_usage.py         # 使用示例
+├── run_asr.bat              # Windows 运行脚本
+├── fix_path_issue.bat       # 路径修复脚本（推荐）
+├── enable_long_path.ps1     # 启用长路径支持（管理员）
+├── requirements.txt         # 依赖列表
+└── README.md                # 本文档
+```
+
+## 快速开始
+
+### Windows 用户注意 ⚠️
+
+如果遇到 **"文件名或扩展名太长"** 错误，请使用以下方法之一：
+
+#### 方法 1：使用 fix_path_issue.bat（推荐）
+```bash
+fix_path_issue.bat your_audio.wav
+```
+
+#### 方法 2：启用 Windows 长路径支持（永久解决）
+1. 右键 PowerShell → 以管理员身份运行
+2. 运行：`enable_long_path.ps1`
+3. 重启电脑
+
+### 1. 激活虚拟环境
+
+```bash
+# Windows CMD
+funasr_env\Scripts\activate.bat
+
+# Linux/Mac
+source funasr_env/bin/activate
+```
+
+### 2. 安装依赖（如果未安装）
+
+```bash
+pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
+pip install funasr modelscope
+```
+
+### 3. 测试识别
+
+```bash
+# 方法 1：使用修复脚本（推荐 Windows 用户）
+fix_path_issue.bat your_audio.wav
+
+# 方法 2：直接运行
+python test_asr.py -f your_audio.wav
+
+# 批量识别目录
+python test_asr.py -d ./audio_files/
+
+# 使用 SenseVoice 模型（多语言）
+python test_asr.py -f your_audio.wav -m SenseVoice
+```
+
+## 代码使用示例
+
+### 基础识别
+
+```python
+from asr_service import ASRService
+
+# 初始化服务
+service = ASRService(model_name="paraformer-zh")
+
+# 识别音频
+sentences = service.recognize("meeting.wav")
+
+# 打印结果
+for sent in sentences:
+    print(f"[{sent.speaker}] {sent.text}")
+    print(f"  时间: {sent.begin_time:.2f}s - {sent.end_time:.2f}s")
+```
+
+### 导出结果
+
+```python
+# 导出为 JSON
+service.export_to_json(sentences, "result.json")
+
+# 导出为 SRT 字幕
+service.export_to_srt(sentences, "result.srt")
+```
+
+## 输出格式
+
+### JSON 格式
+
+```json
+{
+  "total_sentences": 3,
+  "sentences": [
+    {
+      "speaker": "SPEAKER_00",
+      "text": "大家好，今天的会议现在开始。",
+      "begin_time": 0.50,
+      "end_time": 3.20,
+      "duration": 2.70
+    },
+    {
+      "speaker": "SPEAKER_01",
+      "text": "好的，我先汇报一下进度。",
+      "begin_time": 3.50,
+      "end_time": 6.10,
+      "duration": 2.60
+    }
+  ]
+}
+```
+
+### SRT 字幕格式
+
+```srt
+1
+00:00:00,500 --> 00:00:03,200
+[SPEAKER_00] 大家好，今天的会议现在开始。
+
+2
+00:00:03,500 --> 00:00:06,100
+[SPEAKER_01] 好的，我先汇报一下进度。
+```
+
+## 支持的音频格式
+
+- WAV
+- MP3
+- M4A
+- FLAC
+- OGG
+- WMA
+
+## 模型选择
+
+| 模型 | 说明 | 适用场景 |
+|------|------|----------|
+| `paraformer-zh` | 达摩院中文模型（默认） | 中文语音识别，支持说话人分离 |
+| `SenseVoice` | 多语言模型 | 多语言、方言、情感识别 |
+
+## 硬件要求
+
+| 配置 | 说明 |
+|------|------|
+| CPU | 支持，速度较慢 |
+| GPU | 推荐，RTF < 0.01 |
+| 内存 | 4GB+ |
+| 显存 | 2GB+ (GPU 模式) |
+| 磁盘 | 2GB+（模型缓存） |
+
+## 模型下载
+
+首次运行会自动从魔搭社区下载模型到 `models/` 目录：
+- Paraformer: ~500MB
+- VAD: ~100MB
+- 说话人模型: ~100MB
+
+模型下载来源：
+- 魔搭社区：https://modelscope.cn
+- 模型缓存：`./models/`（项目目录下）
+
+## 常见问题
+
+### Q: 文件名太长错误？
+
+**A:** Windows 默认路径长度限制为 260 字符。解决方法：
+1. 使用 `fix_path_issue.bat` 运行（已配置短路径）
+2. 运行 `enable_long_path.ps1` 启用系统长路径支持（需管理员权限+重启）
+
+### Q: 如何准备测试音频？
+
+**A:**
+- 自行录制会议/对话音频
+- AISHELL 开源数据集：https://www.openslr.org/33/
+
+### Q: 支持多人同时说话吗？
+
+**A:** 支持。说话人分离模块会自动区分不同说话人。
+
+### Q: 对噪音环境有什么优化？
+
+**A:** 集成了 FSMN-VAD 语音活动检测，能有效过滤背景噪音。
+
+### Q: 如何切换 CPU/GPU？
+
+**A:**
+```python
+# CPU
+service = ASRService(device="cpu")
+
+# GPU
+service = ASRService(device="cuda")
+
+# 自动选择
+service = ASRService(device="auto")
+```
+
+## 参考链接
+
+- FunASR 官方仓库：https://github.com/alibaba-damo-academy/FunASR
+- 魔搭社区：https://modelscope.cn
+- PyTorch 安装：https://pytorch.org/get-started/locally/
+
+## 许可证
+
+本项目使用 Apache-2.0 许可证
+
+## 运行
+run.bat VID_20251031_132320_019_mono.wav

asr_service.py

@@ -0,0 +1,347 @@
+"""
+FunASR 语音识别服务
+支持：句级时间戳、说话人分离、抗噪
+"""
+
+import os
+import sys
+
+# 解决 Windows 路径长度限制问题
+# 设置模型缓存目录为短路径
+MODEL_CACHE_DIR = os.path.join(os.path.dirname(os.path.abspath(__file__)), "models")
+os.makedirs(MODEL_CACHE_DIR, exist_ok=True)
+os.environ["MODELSCOPE_CACHE"] = MODEL_CACHE_DIR
+os.environ["FUNASR_MODELS_DIR"] = MODEL_CACHE_DIR
+
+# Windows 长路径支持（Windows 10 1607+）
+if sys.platform == "win32":
+    os.environ["PYTHONLEGACYWINDOWSFSENCODING"] = "1"
+
+import json
+from pathlib import Path
+from typing import List, Dict, Union, Optional
+from dataclasses import dataclass
+import warnings
+
+warnings.filterwarnings('ignore')
+
+
+@dataclass
+class Sentence:
+    """识别结果句子"""
+    speaker: str
+    text: str
+    begin_time: float
+    end_time: float
+
+    def to_dict(self) -> Dict:
+        return {
+            "speaker": self.speaker,
+            "text": self.text,
+            "begin_time": round(self.begin_time, 2),
+            "end_time": round(self.end_time, 2),
+            "duration": round(self.end_time - self.begin_time, 2)
+        }
+
+    def __str__(self) -> str:
+        return f"[{self.speaker}] {self.text} ({self.begin_time:.2f}s - {self.end_time:.2f}s)"
+
+
+class ASRService:
+    """
+    语音识别服务
+
+    功能：
+    1. 语音识别（ASR）
+    2. 句级时间戳
+    3. 说话人分离（Speaker Diarization）
+    4. 语音活动检测（VAD）- 抗噪
+    """
+
+    def __init__(
+        self,
+        model_name: str = "paraformer-zh",  # paraformer-zh 或 SenseVoice
+        device: str = "auto",
+        cache_dir: Optional[str] = None
+    ):
+        """
+        初始化 ASR 服务
+
+        Args:
+            model_name: 模型名称
+                - "paraformer-zh": 达摩院 Paraformer 模型（推荐中文）
+                - "SenseVoice": SenseVoice 多语言模型
+            device: 运行设备 ("cpu", "cuda", "auto")
+            cache_dir: 模型缓存目录
+        """
+        self.model_name = model_name
+        self.device = device
+        self.cache_dir = cache_dir or MODEL_CACHE_DIR
+
+        # 确保缓存目录存在
+        os.makedirs(self.cache_dir, exist_ok=True)
+
+        # 处理设备参数
+        self.device = self._get_device(device)
+
+        # 延迟加载模型
+        self._model = None
+
+    def _get_device(self, device: str) -> str:
+        """
+        处理设备参数
+
+        Args:
+            device: 用户指定的设备 ("cpu", "cuda", "auto")
+
+        Returns:
+            str: 实际的设备 ("cpu" 或 "cuda")
+        """
+        import torch
+
+        if device == "auto":
+            # 自动检测 CUDA 是否可用
+            if torch.cuda.is_available():
+                device = "cuda"
+                print(f"检测到 GPU: {torch.cuda.get_device_name(0)}")
+            else:
+                device = "cpu"
+                print("未检测到 GPU，使用 CPU 运行")
+        elif device not in ["cpu", "cuda"]:
+            raise ValueError(f"不支持的设备: {device}，请选择 'cpu', 'cuda' 或 'auto'")
+
+        return device
+
+    def _load_model(self):
+        """懒加载模型"""
+        if self._model is not None:
+            return
+
+        try:
+            from funasr import AutoModel
+        except ImportError:
+            raise ImportError("请安装 FunASR: pip install funasr")
+
+        print(f"正在加载模型: {self.model_name}")
+        print(f"设备: {self.device}")
+        print(f"模型缓存目录: {self.cache_dir}")
+
+        # 模型配置
+        if self.model_name == "paraformer-zh":
+            # Paraformer 中文模型配置（支持时间戳和说话人分离）
+            # 注意：只有以下模型支持时间戳：
+            # - speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch
+            # - speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch
+            self._model = AutoModel(
+                model="iic/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
+                vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch",
+                punc_model="iic/punc_ct-transformer_cn-en-common-vocab471067-large",
+                spk_model="iic/speech_campplus_sv_zh-cn_16k-common",
+                device=self.device,
+                ncpu=4,
+                disable_pbar=True,
+                disable_log=True,
+            )
+        elif self.model_name == "SenseVoice":
+            # SenseVoice 多语言模型配置
+            self._model = AutoModel(
+                model="iic/SenseVoiceSmall",
+                vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch",
+                vad_kwargs={"max_single_segment_time": 30000},
+                device=self.device,
+                disable_pbar=True,
+                disable_log=True,
+            )
+        else:
+            raise ValueError(f"不支持的模型: {self.model_name}")
+
+        print(f"模型加载完成！")
+
+    def recognize(
+        self,
+        audio_path: Union[str, Path],
+        batch_size_s: int = 300,
+        return_raw: bool = False
+    ) -> Union[List[Sentence], Dict]:
+        """
+        识别音频文件
+
+        Args:
+            audio_path: 音频文件路径
+            batch_size_s: 批处理时长（秒）
+            return_raw: 是否返回原始结果
+
+        Returns:
+            List[Sentence]: 识别结果列表（默认）
+            Dict: 原始结果（如果 return_raw=True）
+        """
+        self._load_model()
+
+        audio_path = Path(audio_path)
+        if not audio_path.exists():
+            raise FileNotFoundError(f"音频文件不存在: {audio_path}")
+
+        print(f"正在识别: {audio_path}")
+
+        # 执行识别
+        result = self._model.generate(
+            input=str(audio_path),
+            batch_size_s=batch_size_s,
+            return_raw_text=True,
+            return_spk_res=True,
+        )
+
+        if return_raw:
+            return result
+
+        # 解析结果
+        return self._parse_result(result)
+
+    def _parse_result(self, result: List[Dict]) -> List[Sentence]:
+        """解析识别结果为 Sentence 列表"""
+        sentences = []
+
+        if not result:
+            return sentences
+
+        # FunASR 返回的是列表，取第一个元素
+        res = result[0] if isinstance(result, list) else result
+
+        # 提取句子列表
+        if "sentence_info" in res:
+            # 有说话人分离的情况
+            for sent_info in res["sentence_info"]:
+                sentence = Sentence(
+                    speaker=sent_info.get("speaker", "SPEAKER_00"),
+                    text=sent_info.get("text", "").strip(),
+                    begin_time=sent_info.get("start", 0) / 1000.0,  # ms -> s
+                    end_time=sent_info.get("end", 0) / 1000.0
+                )
+                if sentence.text:
+                    sentences.append(sentence)
+        elif "text" in res:
+            # 纯文本结果（没有时间戳和说话人）
+            sentences.append(Sentence(
+                speaker="SPEAKER_00",
+                text=res["text"].strip(),
+                begin_time=0.0,
+                end_time=0.0
+            ))
+
+        return sentences
+
+    def recognize_batch(
+        self,
+        audio_paths: List[Union[str, Path]],
+        batch_size_s: int = 300
+    ) -> List[List[Sentence]]:
+        """
+        批量识别多个音频文件
+
+        Args:
+            audio_paths: 音频文件路径列表
+            batch_size_s: 批处理时长（秒）
+
+        Returns:
+            List[List[Sentence]]: 每个音频的识别结果
+        """
+        results = []
+        for audio_path in audio_paths:
+            try:
+                result = self.recognize(audio_path, batch_size_s)
+                results.append(result)
+            except Exception as e:
+                print(f"识别失败 [{audio_path}]: {e}")
+                results.append([])
+        return results
+
+    def export_to_json(
+        self,
+        sentences: List[Sentence],
+        output_path: Union[str, Path]
+    ):
+        """
+        导出识别结果为 JSON 文件
+
+        Args:
+            sentences: 识别结果列表
+            output_path: 输出文件路径
+        """
+        output_path = Path(output_path)
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+
+        data = {
+            "total_sentences": len(sentences),
+            "sentences": [s.to_dict() for s in sentences]
+        }
+
+        with open(output_path, "w", encoding="utf-8") as f:
+            json.dump(data, f, ensure_ascii=False, indent=2)
+
+        print(f"结果已保存: {output_path}")
+
+    def export_to_srt(
+        self,
+        sentences: List[Sentence],
+        output_path: Union[str, Path]
+    ):
+        """
+        导出识别结果为 SRT 字幕文件
+
+        Args:
+            sentences: 识别结果列表
+            output_path: 输出文件路径
+        """
+        output_path = Path(output_path)
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+
+        def format_time(seconds: float) -> str:
+            """格式化为 SRT 时间格式"""
+            hours = int(seconds // 3600)
+            minutes = int((seconds % 3600) // 60)
+            secs = int(seconds % 60)
+            millis = int((seconds % 1) * 1000)
+            return f"{hours:02d}:{minutes:02d}:{secs:02d},{millis:03d}"
+
+        with open(output_path, "w", encoding="utf-8") as f:
+            for i, sentence in enumerate(sentences, 1):
+                f.write(f"{i}\n")
+                f.write(f"{format_time(sentence.begin_time)} --> {format_time(sentence.end_time)}\n")
+                f.write(f"[{sentence.speaker}] {sentence.text}\n\n")
+
+        print(f"字幕已保存: {output_path}")
+
+
+# 便捷函数
+def recognize_audio(
+    audio_path: Union[str, Path],
+    model_name: str = "paraformer-zh",
+    device: str = "auto"
+) -> List[Sentence]:
+    """
+    快速识别音频文件
+
+    Args:
+        audio_path: 音频文件路径
+        model_name: 模型名称
+        device: 运行设备
+
+    Returns:
+        List[Sentence]: 识别结果
+    """
+    service = ASRService(model_name=model_name, device=device)
+    return service.recognize(audio_path)
+
+
+if __name__ == "__main__":
+    # 示例用法
+    print("=" * 60)
+    print("FunASR 语音识别服务")
+    print("=" * 60)
+    print("\n支持的音频格式: wav, mp3, m4a, flac 等")
+    print("\n使用方法:")
+    print('  from asr_service import ASRService')
+    print('  service = ASRService()')
+    print('  results = service.recognize("your_audio.wav")')
+    print('  for sent in results:')
+    print('      print(sent)')

enable_long_path.ps1

@@ -0,0 +1,46 @@
+# 启用 Windows 长路径支持（需要管理员权限）
+# 运行后重启电脑生效
+
+Write-Host "========================================" -ForegroundColor Cyan
+Write-Host "启用 Windows 长路径支持" -ForegroundColor Cyan
+Write-Host "========================================" -ForegroundColor Cyan
+Write-Host ""
+
+# 检查是否以管理员身份运行
+if (-NOT ([Security.Principal.WindowsPrincipal][Security.Principal.WindowsIdentity]::GetCurrent()).IsInRole([Security.Principal.WindowsBuiltInRole] "Administrator")) {
+    Write-Host "❌ 请以管理员身份运行 PowerShell 后再执行此脚本" -ForegroundColor Red
+    Write-Host "   右键点击 PowerShell -> 以管理员身份运行" -ForegroundColor Yellow
+    pause
+    exit
+}
+
+# 启用长路径支持
+Write-Host "正在启用长路径支持..." -ForegroundColor Yellow
+try {
+    Set-ItemProperty -Path "HKLM:\SYSTEM\CurrentControlSet\Control\FileSystem" -Name "LongPathsEnabled" -Value 1
+    Write-Host "✅ 注册表修改成功" -ForegroundColor Green
+} catch {
+    Write-Host "❌ 修改失败: $_" -ForegroundColor Red
+    pause
+    exit
+}
+
+# 启用 Python 长路径支持
+Write-Host ""
+Write-Host "Python 长路径环境变量:" -ForegroundColor Yellow
+$envVar = [Environment]::GetEnvironmentVariable("PYTHONLEGACYWINDOWSFSENCODING", "User")
+if ($envVar -eq $null) {
+    [Environment]::SetEnvironmentVariable("PYTHONLEGACYWINDOWSFSENCODING", "1", "User")
+    Write-Host "✅ 已设置 PYTHONLEGACYWINDOWSFSENCODING=1" -ForegroundColor Green
+} else {
+    Write-Host "   已存在: PYTHONLEGACYWINDOWSFSENCODING=$envVar" -ForegroundColor Cyan
+}
+
+Write-Host ""
+Write-Host "========================================" -ForegroundColor Green
+Write-Host "✅ 设置完成！" -ForegroundColor Green
+Write-Host "========================================" -ForegroundColor Green
+Write-Host ""
+Write-Host "注意: 需要重启电脑才能完全生效" -ForegroundColor Yellow
+Write-Host ""
+pause

example_usage.py

@@ -0,0 +1,128 @@
+"""
+FunASR 使用示例
+展示常见的语音识别应用场景
+"""
+
+from asr_service import ASRService, recognize_audio
+
+
+def example_1_basic_recognition():
+    """示例1: 基础识别"""
+    print("=" * 60)
+    print("示例1: 基础语音识别")
+    print("=" * 60)
+
+    # 方式1: 使用便捷函数
+    # results = recognize_audio("meeting.wav")
+
+    # 方式2: 使用服务类（推荐，可复用）
+    service = ASRService(model_name="paraformer-zh")
+    # results = service.recognize("meeting.wav")
+
+    print("代码:")
+    print("  from asr_service import recognize_audio")
+    print("  results = recognize_audio('meeting.wav')")
+    print("  for sent in results:")
+    print("      print(sent)")
+    print()
+    print("输出格式:")
+    print("  [SPEAKER_00] 大家好，今天的会议现在开始。 (0.50s - 3.20s)")
+    print("  [SPEAKER_01] 好的，我先汇报一下进度。 (3.50s - 6.10s)")
+
+
+def example_2_batch_processing():
+    """示例2: 批量处理"""
+    print("\n" + "=" * 60)
+    print("示例2: 批量处理多个音频")
+    print("=" * 60)
+
+    print("代码:")
+    print("  from pathlib import Path")
+    print("  from asr_service import ASRService")
+    print()
+    print("  service = ASRService()")
+    print("  audio_files = list(Path('./audio').glob('*.wav'))")
+    print("  results = service.recognize_batch(audio_files)")
+    print()
+    print("  for audio_path, sentences in zip(audio_files, results):")
+    print("      print(f'{audio_path}: {len(sentences)} 句话')")
+
+
+def example_3_export_results():
+    """示例3: 导出结果"""
+    print("\n" + "=" * 60)
+    print("示例3: 导出识别结果")
+    print("=" * 60)
+
+    print("代码:")
+    print("  service = ASRService()")
+    print("  sentences = service.recognize('meeting.wav')")
+    print()
+    print("  # 导出为 JSON")
+    print("  service.export_to_json(sentences, 'meeting.json')")
+    print()
+    print("  # 导出为 SRT 字幕")
+    print("  service.export_to_srt(sentences, 'meeting.srt')")
+    print()
+    print("JSON 输出示例:")
+    print("""  {
+    "total_sentences": 2,
+    "sentences": [
+      {
+        "speaker": "SPEAKER_00",
+        "text": "大家好",
+        "begin_time": 0.50,
+        "end_time": 3.20,
+        "duration": 2.70
+      }
+    ]
+  }""")
+
+
+def example_4_different_models():
+    """示例4: 选择不同模型"""
+    print("\n" + "=" * 60)
+    print("示例4: 选择不同模型")
+    print("=" * 60)
+
+    print("模型选择:")
+    print()
+    print("1. paraformer-zh (默认)")
+    print("   - 达摩院出品，中文识别精度高")
+    print("   - 支持说话人分离")
+    print("   - 代码: ASRService(model_name='paraformer-zh')")
+    print()
+    print("2. SenseVoice")
+    print("   - 多语言支持（中、英、日、韩等）")
+    print("   - 支持情感识别")
+    print("   - 代码: ASRService(model_name='SenseVoice')")
+
+
+def example_5_hardware_options():
+    """示例5: 硬件选择"""
+    print("\n" + "=" * 60)
+    print("示例5: 选择运行设备")
+    print("=" * 60)
+
+    print("设备选项:")
+    print()
+    print("  # 自动选择 (推荐)")
+    print("  service = ASRService(device='auto')")
+    print()
+    print("  # 使用 GPU")
+    print("  service = ASRService(device='cuda')")
+    print()
+    print("  # 使用 CPU")
+    print("  service = ASRService(device='cpu')")
+
+
+if __name__ == "__main__":
+    example_1_basic_recognition()
+    example_2_batch_processing()
+    example_3_export_results()
+    example_4_different_models()
+    example_5_hardware_options()
+
+    print("\n" + "=" * 60)
+    print("提示: 运行测试请使用: python test_asr.py -f your_audio.wav")
+    print("=" * 60)

requirements.txt

@@ -0,0 +1,7 @@
+funasr>=1.3.0
+modelscope>=1.15.0
+torch>=2.0.0
+torchaudio>=2.0.0
+torchvision>=0.15.0
+transformers>=4.30.0
+numpy>=1.24.0

run.bat

@@ -0,0 +1,36 @@
+@echo off
+chcp 65001 >nul
+echo ========================================
+echo 修复 Windows 路径长度问题
+echo ========================================
+echo.
+
+REM 设置短路径环境变量
+set "MODELSCOPE_CACHE=%~dp0models"
+set "FUNASR_MODELS_DIR=%~dp0models"
+set "PYTHONLEGACYWINDOWSFSENCODING=1"
+
+REM 创建模型目录
+if not exist "models" mkdir models
+
+echo ✅ 环境变量已设置
+echo    MODELSCOPE_CACHE=%MODELSCOPE_CACHE%
+echo    FUNASR_MODELS_DIR=%FUNASR_MODELS_DIR%
+echo.
+
+REM 检查参数
+if "%~1"=="" (
+    echo 使用方法: fix_path_issue.bat [音频文件路径]
+    echo 示例: fix_path_issue.bat meeting.wav
+    pause
+    exit /b 1
+)
+
+echo 🔄 正在运行语音识别...
+echo.
+
+REM 使用虚拟环境的 Python 运行
+funasr_env\Scripts\python.exe test_asr.py -f "%~1"
+
+echo.
+pause

test_asr.py

@@ -0,0 +1,178 @@
+"""
+FunASR 语音识别测试脚本
+测试功能：句级时间戳、说话人分离
+"""
+
+import os
+import sys
+import argparse
+from pathlib import Path
+
+
+def print_banner():
+    """打印欢迎信息"""
+    print("=" * 70)
+    print("                  FunASR 语音识别测试工具")
+    print("=" * 70)
+    print("功能特性:")
+    print("  • 句级时间戳（开始时间 - 结束时间）")
+    print("  • 说话人分离（自动区分不同说话人）")
+    print("  • 抗噪处理（VAD 语音活动检测）")
+    print("  • 支持中文、方言、多语言")
+    print("=" * 70)
+    print()
+
+
+def test_single_audio(audio_path: str, model_name: str = "paraformer-zh"):
+    """测试单个音频文件"""
+    from asr_service import ASRService
+
+    # 检查文件
+    if not os.path.exists(audio_path):
+        print(f"❌ 错误: 文件不存在 - {audio_path}")
+        return
+
+    # 初始化服务
+    print(f"🔄 正在初始化模型: {model_name}")
+    print(f"📝 音频文件: {audio_path}")
+    print("-" * 70)
+
+    service = ASRService(model_name=model_name)
+
+    # 执行识别
+    try:
+        sentences = service.recognize(audio_path)
+    except Exception as e:
+        print(f"❌ 识别失败: {e}")
+        return
+
+    # 显示结果
+    print("\n✅ 识别完成！")
+    print("=" * 70)
+    print(f"共识别出 {len(sentences)} 句话\n")
+
+    for i, sent in enumerate(sentences, 1):
+        print(f"[{i}] {sent}")
+
+    # 导出结果
+    base_name = Path(audio_path).stem
+
+    # 导出 JSON
+    json_path = f"{base_name}_result.json"
+    service.export_to_json(sentences, json_path)
+
+    # 导出 SRT 字幕
+    srt_path = f"{base_name}_result.srt"
+    service.export_to_srt(sentences, srt_path)
+
+    print("\n" + "=" * 70)
+    print("📁 输出文件:")
+    print(f"   • JSON: {json_path}")
+    print(f"   • SRT:  {srt_path}")
+    print("=" * 70)
+
+
+def test_batch(audio_dir: str, model_name: str = "paraformer-zh"):
+    """批量测试目录中的音频文件"""
+    from asr_service import ASRService
+
+    # 支持的音频格式
+    audio_extensions = {".wav", ".mp3", ".m4a", ".flac", ".ogg", ".wma"}
+
+    # 扫描音频文件
+    audio_files = []
+    for ext in audio_extensions:
+        audio_files.extend(Path(audio_dir).glob(f"*{ext}"))
+
+    if not audio_files:
+        print(f"❌ 未找到音频文件（支持格式: {', '.join(audio_extensions)}）")
+        return
+
+    print(f"🔄 找到 {len(audio_files)} 个音频文件")
+    print("-" * 70)
+
+    # 初始化服务
+    service = ASRService(model_name=model_name)
+
+    # 批量识别
+    for audio_path in audio_files:
+        print(f"\n处理: {audio_path.name}")
+        try:
+            sentences = service.recognize(audio_path)
+            print(f"  ✓ 识别出 {len(sentences)} 句话")
+
+            # 导出
+            base_name = audio_path.stem
+            service.export_to_json(sentences, f"{base_name}_result.json")
+        except Exception as e:
+            print(f"  ✗ 失败: {e}")
+
+    print("\n✅ 批量处理完成！")
+
+
+def download_test_audio():
+    """下载测试音频（示例）"""
+    print("📝 请准备测试音频文件")
+    print("支持的格式: wav, mp3, m4a, flac, ogg, wma")
+    print("\n示例音频来源:")
+    print("  • 自行录制会议/对话音频")
+    print("  • AISHELL 开源数据集: https://www.openslr.org/33/")
+    print("  • 魔搭社区示例: https://modelscope.cn/models/iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch")
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="FunASR 语音识别测试工具",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+示例用法:
+  # 识别单个文件
+  python test_asr.py -f your_audio.wav
+
+  # 使用 SenseVoice 模型（多语言）
+  python test_asr.py -f your_audio.wav -m SenseVoice
+
+  # 批量识别目录
+  python test_asr.py -d ./audio_files/
+        """
+    )
+
+    parser.add_argument(
+        "-f", "--file",
+        help="要识别的音频文件路径"
+    )
+    parser.add_argument(
+        "-d", "--directory",
+        help="要批量识别的音频目录"
+    )
+    parser.add_argument(
+        "-m", "--model",
+        default="paraformer-zh",
+        choices=["paraformer-zh", "SenseVoice"],
+        help="选择模型 (默认: paraformer-zh)"
+    )
+    parser.add_argument(
+        "--download-sample",
+        action="store_true",
+        help="显示测试音频下载信息"
+    )
+
+    args = parser.parse_args()
+
+    print_banner()
+
+    if args.download_sample:
+        download_test_audio()
+    elif args.file:
+        test_single_audio(args.file, args.model)
+    elif args.directory:
+        test_batch(args.directory, args.model)
+    else:
+        parser.print_help()
+        print("\n" + "=" * 70)
+        print("提示: 使用 -f 指定音频文件，或 -d 指定音频目录")
+        print("=" * 70)
+
+
+if __name__ == "__main__":
+    main()