[Tools] Buzz – 免费开源的AI语音转文字工具

Buzz 介绍

https://github.com/chidiwilliams/buzz

Buzz是一款基于OpenAI Whisper模型构建的离线语音转文字工具，适用于Windows、macOS和Linux系统。
Buzz能将麦克风输入或音频、视频文件实时转换为文字，支持多种格式导入导出，如TXT、SRT和VTT。
Buzz的转换速度快，准确率高，支持多语言识别，并能将结果翻译成英文。

转换目标

标准答案

Whisper + Tiny 测试

Tiny小模型的识别有瑕疵

WEBVTT

00:00:00.000 --> 00:00:05.440
I want to see the sea.

00:00:05.440 --> 00:00:08.940
It is time to lay eggs.

00:00:08.940 --> 00:00:14.440
It is the first time these young hands have ever laid eggs.

00:00:14.440 --> 00:00:18.000
And for some of them, it hurts.

00:00:18.000 --> 00:00:24.320
Oh, what lovely eggs you've laid!

00:00:24.320 --> 00:00:28.560
The mother hands are very happy.

00:00:28.560 --> 00:00:34.120
Only the young hands Carmella refuses to lay eggs.

00:00:34.120 --> 00:00:36.760
Like eggs, like eggs.

00:00:36.760 --> 00:00:40.560
Always asking us to lay eggs.

00:00:40.560 --> 00:00:44.160
She grumbles unhappily.

00:00:44.160 --> 00:00:50.560
There must be more interesting things to do in our lives.

Whisper + Medium 测试

Medium模型的识别效果蛮好

WEBVTT

00:00:00.000 --> 00:00:05.000
I want to see the sea.

00:00:05.000 --> 00:00:08.000
It is time to lay eggs.

00:00:08.000 --> 00:00:18.000
It is the first time these young hens have ever laid eggs, and for some of them it hurts.

00:00:18.000 --> 00:00:24.000
Oh, what lovely eggs you've laid!

00:00:24.000 --> 00:00:28.000
The mother hens are very happy.

00:00:28.000 --> 00:00:33.000
Only the young hen, Carmela, refuses to lay eggs.

00:00:33.000 --> 00:00:40.000
Lay eggs, lay eggs, always asking us to lay eggs!

00:00:40.000 --> 00:00:43.000
She grumbles unhappily.

00:00:43.000 --> 00:00:49.000
There must be more interesting things to do in our lives.

参考资料

https://ai-bot.cn/buzz/