C#使用whisper.net实现语音识别(语音转文本)

目录

介绍

效果

输出信息

项目

代码

下载


介绍

github地址:https://github.com/sandrohanea/whisper.net

Whisper.net. Speech to text made simple using Whisper Models

模型下载地址:https://huggingface.co/sandrohanea/whisper.net/tree/main/classic

效果

输出信息

whisper_init_from_file_no_state: loading model from 'ggml-small.bin'

whisper_model_load: loading model

whisper_model_load: n_vocab = 51865

whisper_model_load: n_audio_ctx = 1500

whisper_model_load: n_audio_state = 768

whisper_model_load: n_audio_head = 12

whisper_model_load: n_audio_layer = 12

whisper_model_load: n_text_ctx = 448

whisper_model_load: n_text_state = 768

whisper_model_load: n_text_head = 12

whisper_model_load: n_text_layer = 12

whisper_model_load: n_mels = 80

whisper_model_load: ftype = 1

whisper_model_load: qntvr = 0

whisper_model_load: type = 3

whisper_model_load: mem required = 743.00 MB (+ 16.00 MB per decoder)

whisper_model_load: adding 1608 extra tokens

whisper_model_load: model ctx = 464.68 MB

whisper_model_load: model size = 464.44 MB

whisper_init_state: kv self size = 15.75 MB

whisper_init_state: kv cross size = 52.73 MB

00:00:00->00:00:20: 皇鶴楼,崔昊,西人已成皇鶴去,此地空于皇鶴楼,皇鶴一去不复返,白云千载空悠悠。

00:00:20->00:00:39: 青川莉莉汉阳树,方草七七英五周,日暮相关何处事,燕泊江上世人愁。

项目

代码

using System;

using System.Collections.Generic;

using System.ComponentModel;

using System.Data;

using System.Drawing;

using System.IO;

using System.Linq;

using System.Text;

using System.Threading.Tasks;

using System.Windows.Forms;

using Whisper.net;

using static System.Net.Mime.MediaTypeNames;

namespace C_使用whisper.net实现语音转文本

{

public partial class Form1 : Form

{

public Form1()

{

InitializeComponent();

}

string fileFilter = "*.wav|*.wav";

string wavFileName = "";

WhisperFactory whisperFactory;

WhisperProcessor processor;

private async void button2_Click(object sender, EventArgs e)

{

if (wavFileName == "")

{

return;

}

try

{

button2.Enabled = false;

using var fileStream = File.OpenRead(wavFileName);

await foreach (var result in processor.ProcessAsync(fileStream))

{

Console.WriteLine($"{result.Start}->{result.End}: {result.Text}\r\n");

txtResult.Text += $"{result.Start}->{result.End}: {result.Text}\r\n";

}

}

catch (Exception ex)

{

MessageBox.Show(ex.Message);

}

finally

{

button2.Enabled = true;

}

}

private void Form1_Load(object sender, EventArgs e)

{

whisperFactory = WhisperFactory.FromPath("ggml-small.bin");

processor = whisperFactory.CreateBuilder()

.WithLanguage("zh")//.WithLanguage("auto")

.Build();

wavFileName = "085黄鹤楼.wav";

txtFileName.Text = wavFileName;

}

private void button1_Click(object sender, EventArgs e)

{

OpenFileDialog ofd = new OpenFileDialog();

ofd.Filter = fileFilter;

if (ofd.ShowDialog() != DialogResult.OK) return;

txtResult.Text = "";

wavFileName = ofd.FileName;

txtFileName.Text = wavFileName;

}

}

}

复制代码
using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.IO;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Windows.Forms;
using Whisper.net;
using static System.Net.Mime.MediaTypeNames;

namespace C_使用whisper.net实现语音转文本
{
    public partial class Form1 : Form
    {
        public Form1()
        {
            InitializeComponent();
        }

        string fileFilter = "*.wav|*.wav";
        string wavFileName = "";
        WhisperFactory whisperFactory;
        WhisperProcessor processor;
        private async void button2_Click(object sender, EventArgs e)
        {
            if (wavFileName == "")
            {
                return;
            }

            try
            {
                button2.Enabled = false;
                using var fileStream = File.OpenRead(wavFileName);
                await foreach (var result in processor.ProcessAsync(fileStream))
                {
                    Console.WriteLine($"{result.Start}->{result.End}: {result.Text}\r\n");
                    txtResult.Text += $"{result.Start}->{result.End}: {result.Text}\r\n";
                }
            }
            catch (Exception ex)
            {
                MessageBox.Show(ex.Message);
            }
            finally
            {
                button2.Enabled = true;
            }

        }

        private void Form1_Load(object sender, EventArgs e)
        {
            whisperFactory = WhisperFactory.FromPath("ggml-small.bin");

            processor = whisperFactory.CreateBuilder()
               .WithLanguage("zh")//.WithLanguage("auto")
               .Build();

            wavFileName = "085黄鹤楼.wav";
            txtFileName.Text = wavFileName;
        }

        private void button1_Click(object sender, EventArgs e)
        {
            OpenFileDialog ofd = new OpenFileDialog();
            ofd.Filter = fileFilter;
            if (ofd.ShowDialog() != DialogResult.OK) return;

            txtResult.Text = "";

            wavFileName = ofd.FileName;
            txtFileName.Text = wavFileName;
        }
    }
}

下载

源码下载

相关推荐
深圳雨林凯AI7 小时前
商品详情长图的自动生成:版块模板、要素抽取与版式约束的实现思路
人工智能·ai
lisw057 小时前
迁移学习(Transfer Learning):概念、历史、内容、挑战与展望
人工智能·机器学习·迁移学习
炫酷的大高个7 小时前
98% 连跑 20 次只剩 67%,我算了大模型重试账
人工智能·架构·开源
可乐ea7 小时前
智能体的模型列表为什么要从三处来源合成:一次让 GPT-6.1 Sol 选不上的发布延迟
人工智能·gpt·chatgpt·智能体框架·模型列表·多源数据合成·客户端工程
freewebsys7 小时前
【原始POSTMAN】最老一个版本的postman 大家还记得吗?官方已经不支持没有升级了,下载了组件用AI直接升级了下可以使用了!
人工智能
云南新人和数采集团7 小时前
云南靠谱软件开发公司选型指南:以昆明新腾科技为例
大数据·人工智能·科技
发现你走远了8 小时前
05-权限不是开关:工具调用如何在放行、拦截与追问之间做决定
人工智能·agent·claude
AllData公司负责人8 小时前
AI大模型数据开发底座系列说明书:AllData集成DB-GPT开源项目建设 AI 多模态分析底座,涵盖模态解析/智能大屏/MCP 扩展!
大数据·数据库·人工智能·gpt·数据采集·数据接入·模态解析
寂寞的船8 小时前
【Unity知识速递】第13期 Unity 内存泄漏排查实战:从 Memory Profiler 到 Lua 引用泄漏的生产级全链路诊断
unity·junit·c#·游戏引擎·lua
Jeff-Tsai9 小时前
Day 19:定制层的范围与叠加粒度
.net·架构设计·erp·definition-driven