devbox
从前慢现在也慢
这个屌丝很懒，什么也没留下！
关注作者
热门标签
热门文章
当前位置:   article > 正文

C# RAM Stable Diffusion 提示词反推 Onnx Demo

作者：从前慢现在也慢 | 2024-03-14 09:47:38

赞

踩

C# RAM Stable Diffusion 提示词反推 Onnx Demo

目录

介绍

效果

模型信息

项目

代码

下载

C# RAM Stable Diffusion 提示词反推 Onnx Demo

介绍

github地址：GitHub - xinyu1205/recognize-anything: Open-source and strong foundation image recognition models.

Open-source and strong foundation image recognition models.

效果

模型信息

Model Properties
-------------------------
---------------------------------------------------------------

Inputs
-------------------------
name：input
tensor：Float[1, 3, 384, 384]
---------------------------------------------------------------

Outputs
-------------------------
name：output
tensor：Float[1, 4585]
---------------------------------------------------------------

项目

代码

using Microsoft.ML.OnnxRuntime;
using Microsoft.ML.OnnxRuntime.Tensors;
using OpenCvSharp;
using System;
using System.Collections.Generic;
using System.Drawing;
using System.IO;
using System.Linq;
using System.Runtime.InteropServices;
using System.Text;
using System.Windows.Forms;

namespace Onnx_Demo
{
    public partial class Form1 : Form
    {
        public Form1()
        {
            InitializeComponent();
        }

        string fileFilter = "*.*|*.bmp;*.jpg;*.jpeg;*.tiff;*.tiff;*.png";
        string image_path = "";
        DateTime dt1 = DateTime.Now;
        DateTime dt2 = DateTime.Now;
        string model_path;
        Mat image;

        SessionOptions options;
        InferenceSession onnx_session;
        Tensor<float> input_tensor;
        List<NamedOnnxValue> input_container;
        IDisposableReadOnlyCollection<DisposableNamedOnnxValue> result_infer;
        DisposableNamedOnnxValue[] results_onnxvalue;

        Tensor<float> result_tensors;

        StringBuilder sbTags = new StringBuilder();
        StringBuilder sbTagsCN = new StringBuilder();
        StringBuilder sb = new StringBuilder();

        public string[] class_names;

        List<Tag> ltTag = new List<Tag>();

        private void button1_Click(object sender, EventArgs e)
        {
            OpenFileDialog ofd = new OpenFileDialog();
            ofd.Filter = fileFilter;
            if (ofd.ShowDialog() != DialogResult.OK) return;
            pictureBox1.Image = null;
            image_path = ofd.FileName;
            pictureBox1.Image = new Bitmap(image_path);
            textBox1.Text = "";
            image = new Mat(image_path);
        }

        float[] mean = { 0.485f, 0.456f, 0.406f };
        float[] std = { 0.229f, 0.224f, 0.225f };

        public void Normalize(Mat src)
        {
            src.ConvertTo(src, MatType.CV_32FC3, 1.0 / 255);
            Mat[] bgr = src.Split();
            for (int i = 0; i < bgr.Length; ++i)
            {
                bgr[i].ConvertTo(bgr[i], MatType.CV_32FC1, 1 / std[i], (0.0 - mean[i]) / std[i]);
            }
            Cv2.Merge(bgr, src);
            foreach (Mat channel in bgr)
            {
                channel.Dispose();
            }
        }

        public float[] ExtractMat(Mat src)
        {
            OpenCvSharp.Size size = src.Size();
            int channels = src.Channels();
            float[] result = new float[size.Width * size.Height * channels];
            GCHandle resultHandle = default;
            try
            {
                resultHandle = GCHandle.Alloc(result, GCHandleType.Pinned);
                IntPtr resultPtr = resultHandle.AddrOfPinnedObject();
                for (int i = 0; i < channels; ++i)
                {
                    Mat cmat = new Mat(
                       src.Height, src.Width,
                       MatType.CV_32FC1,
                       resultPtr + i * size.Width * size.Height * sizeof(float));

                    Cv2.ExtractChannel(src, cmat, i);
                    cmat.Dispose();
                }
            }
            finally
            {
                resultHandle.Free();
            }
            return result;
        }

        private void button2_Click(object sender, EventArgs e)
        {
            if (image_path == "")
            {
                return;
            }

            button2.Enabled = false;
            textBox1.Text = "";
            sb.Clear();
            sbTagsCN.Clear();
            sbTags.Clear();
            Application.DoEvents();

            image = new Mat(image_path);

            //图片缩放
            Mat resize_image = new Mat();
            Cv2.Resize(image, resize_image, new OpenCvSharp.Size(384, 384));

            Normalize(resize_image);

            var data = ExtractMat(resize_image);

            resize_image.Dispose();
            image.Dispose();

            // 输入Tensor
            input_tensor = new DenseTensor<float>(data, new[] { 1, 3, 384, 384 });

            //将 input_tensor 放入一个输入参数的容器，并指定名称
            input_container.Add(NamedOnnxValue.CreateFromTensor("input", input_tensor));

            dt1 = DateTime.Now;
            //运行 Inference 并获取结果
            result_infer = onnx_session.Run(input_container);
            dt2 = DateTime.Now;

            // 将输出结果转为DisposableNamedOnnxValue数组
            results_onnxvalue = result_infer.ToArray();

            // 读取第一个节点输出并转为Tensor数据
            result_tensors = results_onnxvalue[0].AsTensor<float>();

            var result_array = result_tensors.ToArray();

            double[] scores = new double[result_array.Length];
            for (int i = 0; i < result_array.Length; i++)
            {
                double score = 1 / (1 + Math.Exp(result_array[i] * -1));
                scores[i] = score;
            }
            List<Tag> tags = new List<Tag>(ltTag);

            List<Tag> topTags = new List<Tag>();
            for (int i = 0; i < scores.Length; i++)
            {
                if (scores[i] > tags[i].Threshold)
                {
                    tags[i].Score = scores[i];
                    topTags.Add(tags[i]);
                }
            }
            topTags.OrderByDescending(x => x.Score).ToList();

            foreach (var item in topTags)
            {
                sbTagsCN.Append(item.NameCN + ",");
                sbTags.Append(item.Name + ",");
            }
            sbTagsCN.Length--;
            sbTags.Length--;

            sb.AppendLine("Tags:" + sbTags.ToString());
            sb.AppendLine("标签:" + sbTagsCN.ToString());
            sb.AppendLine("------------------");
            sb.AppendLine("推理耗时:" + (dt2 - dt1).TotalMilliseconds + "ms");
            textBox1.Text = sb.ToString();
            button2.Enabled = true;
        }

        private void Form1_Load(object sender, EventArgs e)
        {
            model_path = "model/ram.onnx";

            // 创建输出会话，用于输出模型读取信息
            options = new SessionOptions();
            options.LogSeverityLevel = OrtLoggingLevel.ORT_LOGGING_LEVEL_INFO;
            options.AppendExecutionProvider_CPU(0);// 设置为CPU上运行
            // 创建推理模型类，读取本地模型文件
            onnx_session = new InferenceSession(model_path, options);//model_path 为onnx模型文件的路径

            // 创建输入容器
            input_container = new List<NamedOnnxValue>();

            image_path = "test_img/1.jpg";
            pictureBox1.Image = new Bitmap(image_path);
            image = new Mat(image_path);

            string[] thresholdLines = File.ReadAllLines("model/ram_tag_list_threshold.txt");
            string[] tagChineseLines = File.ReadAllLines("model/ram_tag_list_chinese.txt");
            string[] tagLines = File.ReadAllLines("model/ram_tag_list.txt");

            for (int i = 0; i < tagLines.Length; i++)
            {
                ltTag.Add(new Tag { NameCN = tagChineseLines[i], Name = tagLines[i], Threshold = double.Parse(thresholdLines[i]) });
            }
        }

    }
}

  1. using Microsoft.ML.OnnxRuntime;
  2. using Microsoft.ML.OnnxRuntime.Tensors;
  3. using OpenCvSharp;
  4. using System;
  5. using System.Collections.Generic;
  6. using System.Drawing;
  7. using System.IO;
  8. using System.Linq;
  9. using System.Runtime.InteropServices;
  10. using System.Text;
  11. using System.Windows.Forms;
  12.  
  13. namespace Onnx_Demo
  14. {
  15.     public partial class Form1 : Form
  16.     {
  17.         public Form1()
  18.         {
  19.             InitializeComponent();
  20.         }
  21.  
  22.         string fileFilter = "*.*|*.bmp;*.jpg;*.jpeg;*.tiff;*.tiff;*.png";
  23.         string image_path = "";
  24.         DateTime dt1 = DateTime.Now;
  25.         DateTime dt2 = DateTime.Now;
  26.         string model_path;
  27.         Mat image;
  28.  
  29.         SessionOptions options;
  30.         InferenceSession onnx_session;
  31.         Tensor<float> input_tensor;
  32.         List<NamedOnnxValue> input_container;
  33.         IDisposableReadOnlyCollection<DisposableNamedOnnxValue> result_infer;
  34.         DisposableNamedOnnxValue[] results_onnxvalue;
  35.  
  36.         Tensor<float> result_tensors;
  37.  
  38.         StringBuilder sbTags = new StringBuilder();
  39.         StringBuilder sbTagsCN = new StringBuilder();
  40.         StringBuilder sb = new StringBuilder();
  41.  
  42.         public string[] class_names;
  43.  
  44.         List<Tag> ltTag = new List<Tag>();
  45.  
  46.         private void button1_Click(object sender, EventArgs e)
  47.         {
  48.             OpenFileDialog ofd = new OpenFileDialog();
  49.             ofd.Filter = fileFilter;
  50.             if (ofd.ShowDialog() != DialogResult.OK) return;
  51.             pictureBox1.Image = null;
  52.             image_path = ofd.FileName;
  53.             pictureBox1.Image = new Bitmap(image_path);
  54.             textBox1.Text = "";
  55.             image = new Mat(image_path);
  56.         }
  57.  
  58.         float[] mean = { 0.485f, 0.456f, 0.406f };
  59.         float[] std = { 0.229f, 0.224f, 0.225f };
  60.  
  61.         public void Normalize(Mat src)
  62.         {
  63.             src.ConvertTo(src, MatType.CV_32FC3, 1.0 / 255);
  64.             Mat[] bgr = src.Split();
  65.             for (int i = 0; i < bgr.Length; ++i)
  66.             {
  67.                 bgr[i].ConvertTo(bgr[i], MatType.CV_32FC1, 1 / std[i], (0.0 - mean[i]) / std[i]);
  68.             }
  69.             Cv2.Merge(bgr, src);
  70.             foreach (Mat channel in bgr)
  71.             {
  72.                 channel.Dispose();
  73.             }
  74.         }
  75.  
  76.         public float[] ExtractMat(Mat src)
  77.         {
  78.             OpenCvSharp.Size size = src.Size();
  79.             int channels = src.Channels();
  80.             float[] result = new float[size.Width * size.Height * channels];
  81.             GCHandle resultHandle = default;
  82.             try
  83.             {
  84.                 resultHandle = GCHandle.Alloc(result, GCHandleType.Pinned);
  85.                 IntPtr resultPtr = resultHandle.AddrOfPinnedObject();
  86.                 for (int i = 0; i < channels; ++i)
  87.                 {
  88.                     Mat cmat = new Mat(
  89.                        src.Height, src.Width,
  90.                        MatType.CV_32FC1,
  91.                        resultPtr + i * size.Width * size.Height * sizeof(float));
  92.  
  93.                     Cv2.ExtractChannel(src, cmat, i);
  94.                     cmat.Dispose();
  95.                 }
  96.             }
  97.             finally
  98.             {
  99.                 resultHandle.Free();
  100.             }
  101.             return result;
  102.         }
  103.  
  104.         private void button2_Click(object sender, EventArgs e)
  105.         {
  106.             if (image_path == "")
  107.             {
  108.                 return;
  109.             }
  110.  
  111.             button2.Enabled = false;
  112.             textBox1.Text = "";
  113.             sb.Clear();
  114.             sbTagsCN.Clear();
  115.             sbTags.Clear();
  116.             Application.DoEvents();
  117.  
  118.             image = new Mat(image_path);
  119.  
  120.             //图片缩放
  121.             Mat resize_image = new Mat();
  122.             Cv2.Resize(image, resize_image, new OpenCvSharp.Size(384, 384));
  123.  
  124.             Normalize(resize_image);
  125.  
  126.             var data = ExtractMat(resize_image);
  127.  
  128.             resize_image.Dispose();
  129.             image.Dispose();
  130.  
  131.             // 输入Tensor
  132.             input_tensor = new DenseTensor<float>(data, new[] { 1, 3, 384, 384 });
  133.  
  134.             //将 input_tensor 放入一个输入参数的容器，并指定名称
  135.             input_container.Add(NamedOnnxValue.CreateFromTensor("input", input_tensor));
  136.  
  137.             dt1 = DateTime.Now;
  138.             //运行 Inference 并获取结果
  139.             result_infer = onnx_session.Run(input_container);
  140.             dt2 = DateTime.Now;
  141.  
  142.             // 将输出结果转为DisposableNamedOnnxValue数组
  143.             results_onnxvalue = result_infer.ToArray();
  144.  
  145.             // 读取第一个节点输出并转为Tensor数据
  146.             result_tensors = results_onnxvalue[0].AsTensor<float>();
  147.  
  148.             var result_array = result_tensors.ToArray();
  149.  
  150.             double[] scores = new double[result_array.Length];
  151.             for (int i = 0; i < result_array.Length; i++)
  152.             {
  153.                 double score = 1 / (1 + Math.Exp(result_array[i] * -1));
  154.                 scores[i] = score;
  155.             }
  156.             List<Tag> tags = new List<Tag>(ltTag);
  157.  
  158.             List<Tag> topTags = new List<Tag>();
  159.             for (int i = 0; i < scores.Length; i++)
  160.             {
  161.                 if (scores[i] > tags[i].Threshold)
  162.                 {
  163.                     tags[i].Score = scores[i];
  164.                     topTags.Add(tags[i]);
  165.                 }
  166.             }
  167.             topTags.OrderByDescending(x => x.Score).ToList();
  168.  
  169.             foreach (var item in topTags)
  170.             {
  171.                 sbTagsCN.Append(item.NameCN + ",");
  172.                 sbTags.Append(item.Name + ",");
  173.             }
  174.             sbTagsCN.Length--;
  175.             sbTags.Length--;
  176.  
  177.             sb.AppendLine("Tags:" + sbTags.ToString());
  178.             sb.AppendLine("标签:" + sbTagsCN.ToString());
  179.             sb.AppendLine("------------------");
  180.             sb.AppendLine("推理耗时:" + (dt2 - dt1).TotalMilliseconds + "ms");
  181.             textBox1.Text = sb.ToString();
  182.             button2.Enabled = true;
  183.         }
  184.  
  185.         private void Form1_Load(object sender, EventArgs e)
  186.         {
  187.             model_path = "model/ram.onnx";
  188.  
  189.             // 创建输出会话，用于输出模型读取信息
  190.             options = new SessionOptions();
  191.             options.LogSeverityLevel = OrtLoggingLevel.ORT_LOGGING_LEVEL_INFO;
  192.             options.AppendExecutionProvider_CPU(0);// 设置为CPU上运行
  193.             // 创建推理模型类，读取本地模型文件
  194.             onnx_session = new InferenceSession(model_path, options);//model_path 为onnx模型文件的路径
  195.  
  196.             // 创建输入容器
  197.             input_container = new List<NamedOnnxValue>();
  198.  
  199.             image_path = "test_img/1.jpg";
  200.             pictureBox1.Image = new Bitmap(image_path);
  201.             image = new Mat(image_path);
  202.  
  203.             string[] thresholdLines = File.ReadAllLines("model/ram_tag_list_threshold.txt");
  204.             string[] tagChineseLines = File.ReadAllLines("model/ram_tag_list_chinese.txt");
  205.             string[] tagLines = File.ReadAllLines("model/ram_tag_list.txt");
  206.  
  207.             for (int i = 0; i < tagLines.Length; i++)
  208.             {
  209.                 ltTag.Add(new Tag { NameCN = tagChineseLines[i], Name = tagLines[i], Threshold = double.Parse(thresholdLines[i]) });
  210.             }
  211.         }
  212.  
  213.     }
  214. }

下载

源码下载（带模型）

模型下载

声明：本文内容由网友自发贡献，不代表【wpsshop博客】立场，版权归原作者所有，本站不承担相应法律责任。如您发现有侵权的内容，请联系我们。转载请注明出处：https://www.wpsshop.cn/w/从前慢现在也慢/article/detail/233054
推荐阅读
相关标签

Copyright © 2003-2013 www.wpsshop.cn 版权所有，并保留所有权利。

  

闽ICP备14008679号