【貓狗數據集】可視化resnet18的輸出

數據集下載地址：

鏈接：https://pan.baidu.com/s/1l1AnBgkAAEhh0vI5_loWKw
提取碼：2xq4

創建數據集：https://www.cnblogs.com/xiximayou/p/12398285.html

讀取數據集：https://www.cnblogs.com/xiximayou/p/12422827.html

進行訓練：https://www.cnblogs.com/xiximayou/p/12448300.html

保存模型並繼續進行訓練：https://www.cnblogs.com/xiximayou/p/12452624.html

加載保存的模型並測試：https://www.cnblogs.com/xiximayou/p/12459499.html

划分驗證集並邊訓練邊驗證：https://www.cnblogs.com/xiximayou/p/12464738.html

使用學習率衰減策略並邊訓練邊測試：https://www.cnblogs.com/xiximayou/p/12468010.html

利用tensorboard可視化訓練和測試過程：https://www.cnblogs.com/xiximayou/p/12482573.html

從命令行接收參數：https://www.cnblogs.com/xiximayou/p/12488662.html

使用top1和top5准確率來衡量模型：https://www.cnblogs.com/xiximayou/p/12489069.html

使用預訓練的resnet18模型：https://www.cnblogs.com/xiximayou/p/12504579.html

計算數據集的平均值和方差：https://www.cnblogs.com/xiximayou/p/12507149.html

讀取數據集的第二種方式：https://www.cnblogs.com/xiximayou/p/12516735.html

對一張張圖像進行預測(而不是測試集)：https://www.cnblogs.com/xiximayou/p/12522690.html

epoch、batchsize、step之間的關系：https://www.cnblogs.com/xiximayou/p/12405485.html

 

最后讀取訓練好的模型，可視化特征圖，至此貓狗數據集系列就完結了，后面准備着手pyorch-ssd訓練自己的數據集（比如是否口罩檢測）。

直接看代碼吧：visual.py

import cv2
import time
import os
import matplotlib.pyplot as plt
import torch
from torch import nn
import torchvision.models as models
import torchvision.transforms as transforms
import numpy as np
import torchvision
import torch.nn as nn
 
savepath=r'results'
if not os.path.exists(savepath):
    os.mkdir(savepath)
 
def load_model():
  model=torchvision.models.resnet18(pretrained=False)
  model.fc = nn.Linear(model.fc.in_features,2,bias=False)
  save_path="/content/drive/My Drive/colab notebooks/output/resnet18_best.t7" 
  checkpoint = torch.load(save_path,map_location=lambda storage, loc: storage)
  model.load_state_dict(checkpoint['model'])
  return model 

def draw_features(width,height,x,savename):
    tic=time.time()
    fig = plt.figure(figsize=(16, 16))
    fig.subplots_adjust(left=0.05, right=0.95, bottom=0.05, top=0.95, wspace=0.05, hspace=0.05)
    for i in range(width*height):
        plt.subplot(height,width, i + 1)
        plt.axis('off')
        img = x[0, i, :, :]
        pmin = np.min(img)
        pmax = np.max(img)
        img = ((img - pmin) / (pmax - pmin + 0.000001))*255  #float在[0，1]之間，轉換成0-255
        img=img.astype(np.uint8)  #轉成unit8
        img=cv2.applyColorMap(img, cv2.COLORMAP_JET) #生成heat map
        img = img[:, :, ::-1]#注意cv2（BGR）和matplotlib(RGB)通道是相反的
        plt.imshow(img)
        #print("{}/{}".format(i+1,width*height))
    fig.savefig(savename, dpi=100)
    fig.clf()
    plt.close()
    print("time:{}".format(time.time()-tic))
 
 
class ft_net(nn.Module):
 
    def __init__(self):
        super(ft_net, self).__init__()
        model=load_model()
        self.model = model
 
    def forward(self, x):
        if True: # draw features or not
            x = self.model.conv1(x)
            draw_features(8,8,x.cpu().numpy(),"{}/f1_conv1.png".format(savepath))
 
            x = self.model.bn1(x)
            draw_features(8, 8, x.cpu().numpy(),"{}/f2_bn1.png".format(savepath))
 
            x = self.model.relu(x)
            draw_features(8, 8, x.cpu().numpy(), "{}/f3_relu.png".format(savepath))
 
            x = self.model.maxpool(x)
            draw_features(8, 8, x.cpu().numpy(), "{}/f4_maxpool.png".format(savepath))
 
            x = self.model.layer1(x)
            draw_features(8, 8, x.cpu().numpy(), "{}/f5_layer1.png".format(savepath))
 
            x = self.model.layer2(x)
            draw_features(8, 16, x.cpu().numpy(), "{}/f6_layer2.png".format(savepath))
 
            x = self.model.layer3(x)
            draw_features(16, 16, x.cpu().numpy(), "{}/f7_layer3.png".format(savepath))
 
            x = self.model.layer4(x)
            draw_features(16, 32, x.cpu().numpy(), "{}/f8_layer4.png".format(savepath))
            #draw_features(16, 32, x.cpu().numpy()[:, 0:1024, :, :], "{}/f8_layer4_1.png".format(savepath))
            #draw_features(16, 32, x.cpu().numpy()[:, 1024:2048, :, :], "{}/f8_layer4_2.png".format(savepath))
 
            x = self.model.avgpool(x)
            #plt.plot(np.linspace(1, 2048, 2048), x.cpu().numpy()[0, :, 0, 0])
            plt.plot(np.linspace(1, 512, 512), x.cpu().numpy()[0, :, 0, 0])
            plt.savefig("{}/f9_avgpool.png".format(savepath))
            plt.clf()
            plt.close()
 
            x = x.view(x.size(0), -1)
            x = self.model.fc(x)
            #plt.plot(np.linspace(1, 1000, 1000), x.cpu().numpy()[0, :])
            plt.plot(np.linspace(1, 2, 2), x.cpu().numpy()[0, :])
            plt.savefig("{}/f10_fc.png".format(savepath))
            plt.clf()
            plt.close()
        else :
            x = self.model.conv1(x)
            x = self.model.bn1(x)
            x = self.model.relu(x)
            x = self.model.maxpool(x)
            x = self.model.layer1(x)
            x = self.model.layer2(x)
            x = self.model.layer3(x)
            x = self.model.layer4(x)
            x = self.model.avgpool(x)
            x = x.view(x.size(0), -1)
            x = self.model.fc(x)
 
        return x
 
 
model=ft_net()
 
# pretrained_dict = resnet50.state_dict()
# pretrained_dict = {k: v for k, v in pretrained_dict.items() if k in model_dict}
# model_dict.update(pretrained_dict)
# net.load_state_dict(model_dict)
model.eval()
img=cv2.imread('/content/drive/My Drive/colab notebooks/image/cat7.jpg')
img=cv2.resize(img,(224,224))
img=cv2.cvtColor(img,cv2.COLOR_BGR2RGB)
transform = transforms.Compose(
    [transforms.ToTensor(),
     transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))])
img=transform(img)
img=img.unsqueeze(0)
 
with torch.no_grad():
    start=time.time()
    out=model(img)
    print("total time:{}".format(time.time()-start))
    result=out.cpu().numpy()
    # ind=np.argmax(out.cpu().numpy())
    ind=np.argsort(result,axis=1)
    """
    for i in range(5):
        print("predict:top {} = cls {} : score {}".format(i+1,ind[0,1000-i-1],result[0,1000-i-1]))
    """
    for i in range(2):
        print("predict:top {} = cls {} : score {}".format(i+1,ind[0,2-i-1],result[0,2-i-1]))
    print("done")

說明：需要注意的地方

• 在draw_features()中的前兩個參數的乘積必須為該層輸出的通道數目的大小。
• 在GPU上訓練的模型要轉換成CPU模式。
• 輸入的圖像轉換成測試的格式：圖像大小、維度[batchsize,C,H,W]
• 要注意我們的類別是兩類：貓和狗

運行：

輸出文件夾：

原始圖片：

查看每一個文件中的圖像：只截取部分

f1_conv1.png 

f2_bn1.png

f3_relu.png 

f4_maxpool.png

f5_layer1.png

f6_layer2.png

f7_layer3.png

f8_layer4.png

f9_avgpool.png

f10_fc.png

橫軸是類別編號，縱軸是評分。最后一個圖咋好像不太對勁。。

參考了：https://blog.csdn.net/weixin_40500230/article/details/93845890

其實pytorch有一個可視化庫很全面，奈何整了半天沒整成功，有機會再試試了。 

https://github.com/utkuozbulak/pytorch-cnn-visualizations/