feat(kb-workflow): 切换为 Whisper 语音识别

- 将步骤5从 OCR 改为 Whisper 语音识别
- 更新 kb-whisper 支持自动音频转换
- 默认使用中文语言 (zh) 和 base 模型
- 更新输出格式为 {platform}_{id}.txt/.srt/.json
- 添加 FunASR 作为备用方案
This commit is contained in:
2026-02-13 09:04:27 +08:00
parent 9102a973cd
commit 29dd236bb9
7 changed files with 844 additions and 155 deletions
+193 -138
View File
@@ -2,6 +2,11 @@
"""
KB Whisper - 本地语音识别工具
基于 Whisper.cpp 的音视频转文字
功能:
- 自动将视频/音频转换为 16kHz WAV 格式
- 支持指定语言(默认中文)
- 输出标准 JSON 格式供 KB Workflow 使用
"""
import argparse
@@ -9,6 +14,7 @@ import json
import os
import subprocess
import sys
import tempfile
from pathlib import Path
import re
@@ -22,7 +28,7 @@ def parse_arguments():
示例:
python whisper.py --input video.mp4
python whisper.py --input video.mp4 --model base
python whisper.py --input audio.mp3 --output ./output --background
python whisper.py --input video.mp4 --output ./output --background
"""
)
parser.add_argument(
@@ -32,9 +38,9 @@ def parse_arguments():
)
parser.add_argument(
'--model',
default='tiny',
default='base',
choices=['tiny', 'base', 'small', 'medium', 'large'],
help='Whisper 模型类型 (默认: tiny)'
help='Whisper 模型类型 (默认: base)'
)
parser.add_argument(
'--output',
@@ -47,7 +53,8 @@ def parse_arguments():
)
parser.add_argument(
'--language',
help='指定语言 (默认: 自动检测),例如: en, zh, ja'
default='zh',
help='指定语言 (默认: zh),例如: en, zh, ja'
)
return parser.parse_args()
@@ -92,119 +99,184 @@ def check_model(whisper_path, model):
return model_path
def run_whisper_cpp(whisper_path, input_file, model, output_dir, language=None):
"""运行 whisper.cpp 命令行工具"""
def convert_to_wav(input_path, output_dir):
"""将音视频转换为 16kHz WAV 格式
Whisper 需要 WAV 格式的输入文件。
"""
input_path = os.path.abspath(input_path)
# 如果已经是 WAV 且采样率正确,直接返回
if input_path.lower().endswith('.wav'):
try:
result = subprocess.run(
['ffprobe', '-v', 'error', '-show_entries',
'stream=sample_rate', '-of', 'default=noprint_wrappers=1:nokey=1',
input_path],
capture_output=True, text=True
)
if result.returncode == 0 and int(result.stdout.strip()) == 16000:
print(f"输入文件已是 16kHz WAV 格式,直接使用")
return input_path
except Exception:
pass
# 转换
output_path = os.path.join(output_dir, f"{Path(input_path).stem}_audio.wav")
print(f"转换音频格式: {input_path} -> {output_path}")
try:
subprocess.run([
'ffmpeg', '-y', '-i', input_path,
'-acodec', 'pcm_s16le',
'-ar', '16000',
'-ac', '1',
'-loglevel', 'error',
output_path
], check=True, capture_output=True)
print("音频转换完成")
return output_path
except subprocess.CalledProcessError as e:
print(f"音频转换失败: {e}")
raise RuntimeError(f"无法转换音频格式: {e}")
except FileNotFoundError:
raise RuntimeError("未找到 ffmpeg,请安装: sudo apt-get install ffmpeg")
def run_whisper(whisper_path, audio_path, model, output_dir, language='zh'):
"""运行 Whisper 语音识别
Args:
whisper_path: whisper.cpp 目录
audio_path: WAV 音频文件路径
model: 模型名称
output_dir: 输出目录
language: 语言代码
"""
# 输出文件名(使用原始输入文件名,不是 audio.wav)
# 从 output_dir 和 audio_path 推断原始文件名
original_stem = Path(audio_path).stem.replace('_audio', '')
output_basename = original_stem
# 构建命令
cmd = [
os.path.join(whisper_path, 'main'),
os.path.join(whisper_path, 'whisper-cli'),
'-m', os.path.join(whisper_path, 'models', f'ggml-{model}.bin'),
'-f', input_file,
'-of', os.path.join(output_dir, Path(input_file).stem),
'-otxt', # 输出 txt
'-osrt', # 输出 srt
'-oj', # 输出 json
'-f', audio_path,
'-of', os.path.join(output_dir, output_basename),
'-l', language,
'--output-txt',
'--output-srt',
'--output-json',
]
# 添加语言参数(如果指定)
if language:
cmd.extend(['-l', language])
print(f"执行命令: {' '.join(cmd)}")
# 运行命令
# 运行
result = subprocess.run(
cmd,
capture_output=True,
text=True,
cwd=whisper_path
cwd=whisper_path,
timeout=3600 # 1小时超时
)
if result.returncode != 0:
raise RuntimeError(
f"Whisper.cpp 执行失败:\n"
f"标准输出: {result.stdout}\n"
f"错误输出: {result.stderr}"
)
# 检查是否有警告但仍成功
if "error" in result.stderr.lower() and not os.path.exists(
os.path.join(output_dir, f"{output_basename}.txt")
):
raise RuntimeError(
f"Whisper 执行失败:\n"
f"标准输出: {result.stdout}\n"
f"错误输出: {result.stderr}"
)
return result
def parse_whisper_json(json_file):
"""解析 whisper.cpp 输出的 JSON 文件"""
try:
with open(json_file, 'r', encoding='utf-8') as f:
data = json.load(f)
return data
except json.JSONDecodeError:
# whisper.cpp 的 JSON 格式可能不同,尝试解析标准格式
return None
def parse_srt_for_duration(srt_file):
"""从 SRT 文件解析时长"""
if not os.path.exists(srt_file):
return 0
with open(srt_file, 'r', encoding='utf-8') as f:
content = f.read()
# 找到最后一个时间戳
pattern = r'(\d{2}:\d{2}:\d{2}),\d{3} --> \d{2}:\d{2}:\d{2},\d{3}'
matches = re.findall(pattern, content)
if not matches:
return 0
# 解析最后一个时间戳
last_time = matches[-1]
h, m, s = last_time.split(':')
return int(h) * 3600 + int(m) * 60 + int(s)
def create_standard_json(input_file, output_dir, srt_file, txt_file, model):
"""创建标准格式的 JSON 输出"""
basename = Path(input_file).stem
def create_workflow_json(input_file, output_dir, model, language, duration):
"""创建标准格式的 JSON 输出,供 KB Workflow 使用"""
input_path = Path(input_file)
basename = input_path.stem
# 解析 SRT 文件获取时间戳
txt_file = os.path.join(output_dir, f"{basename}.txt")
srt_file = os.path.join(output_dir, f"{basename}.srt")
json_file = os.path.join(output_dir, f"{basename}.json")
# 读取文本内容
text_content = ""
if os.path.exists(txt_file):
with open(txt_file, 'r', encoding='utf-8') as f:
text_content = f.read().strip()
# 解析 SRT 获取分段
segments = []
if os.path.exists(srt_file):
with open(srt_file, 'r', encoding='utf-8') as f:
content = f.read()
# 解析 SRT 格式
# SRT 格式: 序号 -> 时间 -> 文本
pattern = r'(\d+)\n(\d{2}:\d{2}:\d{2},\d{3}) --> (\d{2}:\d{2}:\d{2},\d{3})\n(.+?)(?=\n\n|\Z)'
matches = re.findall(pattern, content, re.DOTALL)
for idx, (num, start, end, text) in enumerate(matches):
# 转换时间戳
def parse_time(t):
h, m, s_ms = t.split(':')
s, ms = s_ms.split(',')
return int(h) * 3600 + int(m) * 60 + int(s) + int(ms) / 1000
def parse_time(t):
h, m, s_ms = t.split(':')
s, ms = s_ms.split(',')
return float(h) * 3600 + float(m) * 60 + float(s) + float(ms) / 1000
for idx, (num, start, end, text) in enumerate(matches):
segments.append({
"id": idx,
"start": parse_time(start),
"end": parse_time(end),
"start": round(parse_time(start), 2),
"end": round(parse_time(end), 2),
"text": text.strip()
})
# 读取纯文本
txt_content = ""
if os.path.exists(txt_file):
with open(txt_file, 'r', encoding='utf-8') as f:
txt_content = f.read().strip()
# 获取音频时长(使用 ffprobe)
duration = 0
try:
result = subprocess.run(
['ffprobe', '-v', 'error', '-show_entries', 'format=duration',
'-of', 'default=noprint_wrappers=1:nokey=1', input_file],
capture_output=True,
text=True
)
if result.returncode == 0:
duration = float(result.stdout.strip())
except FileNotFoundError:
pass
# 构建标准 JSON
json_data = {
# 构建 JSON
workflow_json = {
"success": True,
"input_file": os.path.basename(input_file),
"input_path": input_file,
"tool": "whisper",
"model": model,
"language": language,
"input_file": str(input_path),
"output_files": {
"json": f"{basename}.json",
"txt": f"{basename}.txt",
"srt": f"{basename}.srt",
"txt": f"{basename}.txt"
"json": f"{basename}.json"
},
"duration_seconds": duration,
"segments": segments
"segments": segments,
"text": text_content
}
return json_data
# 写入 JSON
with open(json_file, 'w', encoding='utf-8') as f:
json.dump(workflow_json, f, ensure_ascii=False, indent=2)
return workflow_json
def main():
@@ -228,24 +300,22 @@ def main():
# 获取 whisper.cpp 路径
try:
whisper_path = get_whisper_cpp_path()
print(f"Whisper.cpp 路径: {whisper_path}")
print(f"✓ Whisper.cpp 路径: {whisper_path}")
except FileNotFoundError as e:
print(f"错误: {e}", file=sys.stderr)
print(f"✗ 错误: {e}", file=sys.stderr)
sys.exit(1)
# 检查模型
try:
check_model(whisper_path, args.model)
print(f"使用模型: {args.model}")
print(f"✓ 使用模型: {args.model}")
except FileNotFoundError as e:
print(f"错误: {e}", file=sys.stderr)
print(f"✗ 错误: {e}", file=sys.stderr)
sys.exit(1)
# 后台执行
if args.background:
print("启动后台处理...")
# 使用 nohup 在后台运行
log_file = os.path.join(output_dir, f"{Path(input_file).stem}_whisper.log")
with open(log_file, 'w') as log:
@@ -253,91 +323,76 @@ def main():
[sys.executable, __file__,
'--input', input_file,
'--model', args.model,
'--output', output_dir] +
(['--language', args.language] if args.language else []),
'--output', output_dir,
'--language', args.language],
stdout=log,
stderr=log,
start_new_session=True
)
print(f"后台任务已启动,日志文件: {log_file}")
print(f"输出文件将生成到: {output_dir}")
print(f"✓ 后台任务已启动")
print(f" 日志: {log_file}")
print(f" 输出: {output_dir}")
# 返回任务信息
result = {
"success": True,
"background": True,
"status": "running",
"log_file": log_file,
"output_dir": output_dir,
"message": "后台处理中,请稍候检查输出文件"
"output_dir": output_dir
}
print(json.dumps(result, ensure_ascii=False, indent=2))
return
# 前台执行
print(f"处理文件: {input_file}")
print(f"输出目录: {output_dir}")
print(f"\n📁 处理文件: {input_file}")
print(f"📁 输出目录: {output_dir}")
try:
# 运行 whisper.cpp
result = run_whisper_cpp(
whisper_path,
input_file,
args.model,
output_dir,
args.language
)
# 1. 转换为 WAV
print(f"\n🔄 步骤1: 转换音频格式...")
audio_path = convert_to_wav(input_file, output_dir)
# 输出文件路径
# 2. 运行 Whisper
print(f"\n🎙️ 步骤2: 语音识别 (语言: {args.language})...")
run_whisper(whisper_path, audio_path, args.model, output_dir, args.language)
# 3. 清理临时 WAV 文件
if audio_path != input_file and audio_path.endswith('_audio.wav'):
try:
os.remove(audio_path)
print("✓ 清理临时文件")
except Exception:
pass
# 4. 获取时长并创建 JSON
print(f"\n📝 步骤3: 生成输出文件...")
basename = Path(input_file).stem
json_file = os.path.join(output_dir, f"{basename}.json")
srt_file = os.path.join(output_dir, f"{basename}.srt")
txt_file = os.path.join(output_dir, f"{basename}.txt")
duration = parse_srt_for_duration(srt_file)
# 检查输出文件
output_files = {}
if os.path.exists(srt_file):
output_files['srt'] = srt_file
if os.path.exists(txt_file):
output_files['txt'] = txt_file
# 创建标准 JSON 格式
standard_json = create_standard_json(
input_file,
output_dir,
srt_file,
txt_file,
args.model
workflow_json = create_workflow_json(
input_file, output_dir, args.model, args.language, duration
)
# 写入标准 JSON
with open(json_file, 'w', encoding='utf-8') as f:
json.dump(standard_json, f, ensure_ascii=False, indent=2)
output_files['json'] = json_file
# 输出结果
result = {
"success": True,
"input_file": input_file,
"model": args.model,
"output_files": {
k: os.path.basename(v) for k, v in output_files.items()
},
"output_dir": output_dir,
"total_segments": len(standard_json.get('segments', [])),
"duration": standard_json.get('duration_seconds', 0)
}
print(f"\n✅ 处理完成!")
print(f" 时长: {duration} 秒")
print(f" 分段数: {len(workflow_json['segments'])}")
print(f"\n📄 输出文件:")
for fmt in ['txt', 'srt', 'json']:
fpath = os.path.join(output_dir, f"{basename}.{fmt}")
if os.path.exists(fpath):
size = os.path.getsize(fpath)
print(f" • {fmt.upper()}: {fpath} ({size:,} bytes)")
print("\n✓ 处理完成!")
print(f"\n输出文件:")
for fmt, fname in output_files.items():
print(f" - {fname.upper()}: {fname}")
print(f"\n完整结果:")
print(json.dumps(result, ensure_ascii=False, indent=2))
print(f"\n📋 Workflow JSON:")
print(json.dumps({
"success": workflow_json["success"],
"tool": "whisper",
"text_length": len(workflow_json["text"]),
"segments_count": len(workflow_json["segments"])
}, ensure_ascii=False, indent=2))
except Exception as e:
print(f"\n✗ 错误: {e}", file=sys.stderr)