视频自动字幕

复制import whisper

# 加载模型（首次运行自动下载，约 1.5GB）
model = whisper.load_model("base")

# 转写视频音频，返回带时间戳的段落
result = model.transcribe("demo.mp4", language="zh")

# 生成 SRT 格式字幕
srt_lines = []
for i, seg in enumerate(result["segments"], start=1):
    start = seg["start"]
    end = seg["end"]
    # 时间戳格式：00:00:00,000
    srt_lines.append(f"{i}")
    srt_lines.append(
        f"{int(start//3600):02d}:{int(start%3600//60):02d}:{start%60:06.3f}".replace('.', ',') +
        " --> " +
        f"{int(end//3600):02d}:{int(end%3600//60):02d}:{end%60:06.3f}".replace('.', ',')
    )
    srt_lines.append(seg["text"].strip())
    srt_lines.append("")

with open("output.srt", "w", encoding="utf-8") as f:
    f.write("\n".join(srt_lines))

print("字幕已保存为 output.srt")

复制package main

import (
	"fmt"
	"os"
	"os/exec"
)

// 使用 whisper.cpp 命令行工具（需预先编译 whisper-cli）
func main() {
	// 调用 whisper-cli 处理 demo.mp4，输出 SRT
	cmd := exec.Command("whisper-cli",
		"-m", "models/ggml-base.bin",
		"-f", "demo.mp4",
		"-l", "zh",
		"-osrt",
		"-of", "output",
	)
	cmd.Stdout = os.Stdout
	cmd.Stderr = os.Stderr

	if err := cmd.Run(); err != nil {
		fmt.Fprintf(os.Stderr, "whisper 执行失败: %v\n", err)
		os.Exit(1)
	}

	fmt.Println("字幕已生成: output.srt")
}

复制// 浏览器端使用 transformers.js + Web Audio API
// 注意：需在支持 WebGPU 的浏览器中运行
import { pipeline } from '@xenova/transformers';

async function generateSubtitles(videoFile) {
  // 初始化语音识别 pipeline
  const transcriber = await pipeline('automatic-speech-recognition', 'Xenova/whisper-base');

  // 读取视频文件并提取音频（简化：假设已有 AudioBuffer）
  const audioCtx = new AudioContext();
  const arrayBuffer = await videoFile.arrayBuffer();
  const audioBuffer = await audioCtx.decodeAudioData(arrayBuffer);

  // 转写
  const result = await transcriber(audioBuffer, {
    language: 'zh',
    return_timestamps: true,
  });

  // 组装 SRT
  let srt = '';
  result.chunks.forEach((chunk, i) => {
    const start = chunk.timestamp[0];
    const end = chunk.timestamp[1];
    const fmtTime = (t) => {
      const h = Math.floor(t / 3600);
      const m = Math.floor((t % 3600) / 60);
      const s = (t % 60).toFixed(3).replace('.', ',');
      return `${String(h).padStart(2, '0')}:${String(m).padStart(2, '0')}:${s}`;
    };
    srt += `${i + 1}\n`;
    srt += `${fmtTime(start)} --> ${fmtTime(end)}\n`;
    srt += `${chunk.text.trim()}\n\n`;
  });

  // 下载 SRT 文件
  const blob = new Blob([srt], { type: 'text/plain;charset=utf-8' });
  const url = URL.createObjectURL(blob);
  const a = document.createElement('a');
  a.href = url;
  a.download = 'subtitle.srt';
  a.click();
  URL.revokeObjectURL(url);
}

// 使用示例：
// document.getElementById('videoInput').addEventListener('change', (e) => {
//   generateSubtitles(e.target.files[0]);
// });