可视化特征图

一副图像输入网络中，每一层都会生成特征图，将指定层的特征图进行可视化，可以方便我们直观去理解每层网络提取出的特征是什么。

 

主要分为三个步骤：

1.将输入图片转成tensor形式[1,c,h,w], 其中1代表batchsize为1

2.创建模型，并且指定需要可视化的层

3.保存特征图的每一层，转成伪彩图进行保存

from torchvision import datasets, models, transforms
import torch.nn as nn
import numpy as np
from PIL import Image
import torch
import cv2
import scipy.misc
import matplotlib.pyplot as plt

class ResNet(nn.Module):
    def __init__(self, num_classes=14):   # num_classes，此处为 二分类值为2
        super(ResNet, self).__init__()
        net = models.resnet50(pretrained=True)   # 从预训练模型加载VGG16网络参数
        self.feature = nn.Sequential(*list(net.children())[:-2])

    def forward(self, x):
        x = self.feature(x)
        return x

def get_k_layer_feature_map(model_layer, k, x):
    with torch.no_grad():
        for index, layer in enumerate(model_layer):#model的第一个Sequential()是有多层，所以遍历
            x = layer(x)#torch.Size([1, 64, 55, 55])生成了64个通道
            if k == index:
                return x

def get_image_info(image_dir):
    image_info = Image.open(image_dir).convert('RGB')#是一幅图片
    # 数据预处理方法
    image_transform = transforms.Compose([
        transforms.Resize(256),
        transforms.CenterCrop(224),
        transforms.ToTensor(),
        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
    ])
    image_info = image_transform(image_info)#torch.Size([3, 224, 224])
    image_info = image_info.unsqueeze(0)#torch.Size([1, 3, 224, 224])因为model的输入要求是4维，所以变成4维
    return image_info#变成tensor数据


#  可视化特征图
def show_feature_map(feature_map):  # feature_map=torch.Size([1, 64, 55, 55]),feature_map[0].shape=torch.Size([64, 55, 55])
    # feature_map[2].shape     out of bounds
    feature_map = feature_map.detach().numpy().squeeze()  # 压缩成torch.Size([64, 55, 55])
    feature_map_num = feature_map.shape[0]  # 返回通道数

    for index in range(feature_map_num):  # 通过遍历的方式，将64个通道的tensor拿出
        feature=feature_map[index]
        feature = np.asarray(feature* 255, dtype=np.uint8)
        feature=cv2.resize(feature,(224,224),interpolation =  cv2.INTER_NEAREST) #改变特征呢图尺寸
        feature = cv2.applyColorMap(feature, cv2.COLORMAP_JET) #变成伪彩图
        cv2.imwrite('/home/cai/project/Catheter/Data/feature_map_save/channel_{}.png'.format(str(index)), feature)

#1.将输入图片转成tensor的形式
image_info=get_image_info("/home/cai/project/Catheter/Data/1.png")

#2.指定可视化第一层feature,并且将特征图取出
my_model=ResNet()
model_layer= list(my_model.children())
model_layer=model_layer[0]#这里选择model的第一个Sequential()
feature_map = get_k_layer_feature_map(model_layer, 0, image_info)

#3.将特征图逐层保存
show_feature_map(feature_map)