描述

实践环境

百度AI实训平台-AI Studio、PaddlePaddle2.0.0

深度神经网络

DNN
深度神经网络（Deep Neural Networks，简称DNN）是深度学习的基础，其结构为input、hidden（可有多层）、output，每层均为全连接。

数据集介绍

数据集文件名为archive_train.zip,archive_test.zip。

该数据集包含25个类别不同宝石的图像。

这些类别已经分为训练和测试数据。

图像大小不一，格式为.jpeg。

数据准备

导包

#导入需要的包
import os
import zipfile
import random
import json
import cv2
import numpy as np
from PIL import Image
import paddle
import matplotlib.pyplot as plt
from paddle.io import Dataset

参数配置

train_parameters = {
    "input_size": [3, 224, 224],                           #输入图片的shape
    "class_dim": 25,                                     #分类数
    "src_path":"data/data55032/archive_train.zip",       #原始数据集路径
    "target_path":"/home/aistudio/data/dataset",        #要解压的路径 
    "train_list_path": "./train.txt",              #train_data.txt路径
    "eval_list_path": "./eval.txt",                  #eval_data.txt路径
    "label_dict":{},                                    #标签字典
    "readme_path": "/home/aistudio/data/readme.json",   #readme.json路径
    "num_epochs": 10,                                    #训练轮数
    "train_batch_size": 16,                             #批次的大小
    "learning_strategy": {                              #优化函数相关的配置
        "lr": 0.0005                                     #超参数学习率
    } 
}

解压数据集

def unzip_data(src_path,target_path):

    '''
    解压原始数据集，将src_path路径下的zip包解压至data/dataset目录下
    '''

    if(not os.path.isdir(target_path)):    
        z = zipfile.ZipFile(src_path, 'r')
        z.extractall(path=target_path)
        z.close()
    else:
        print("文件已解压")

定义数据列表

def get_data_list(target_path,train_list_path,eval_list_path):
    '''
    生成数据列表
    '''
    #存放所有类别的信息
    class_detail = []
    #获取所有类别保存的文件夹名称
    data_list_path=target_path
    class_dirs = os.listdir(data_list_path)
    if '__MACOSX' in class_dirs:
        class_dirs.remove('__MACOSX')
    # #总的图像数量
    all_class_images = 0
    # #存放类别标签
    class_label=0
    # #存放类别数目
    class_dim = 0
    # #存储要写进eval.txt和train.txt中的内容
    trainer_list=[]
    eval_list=[]
    #读取每个类别
    for class_dir in class_dirs:
        if class_dir != ".DS_Store":
            class_dim += 1
            #每个类别的信息
            class_detail_list = {}
            eval_sum = 0
            trainer_sum = 0
            #统计每个类别有多少张图片
            class_sum = 0
            #获取类别路径 
            path = os.path.join(data_list_path,class_dir)
            # 获取所有图片
            img_paths = os.listdir(path)
            for img_path in img_paths:                                  # 遍历文件夹下的每个图片
                if img_path =='.DS_Store':
                    continue
                name_path = os.path.join(path,img_path)                       # 每张图片的路径
                if class_sum % 15 == 0:                                 # 每10张图片取一个做验证数据
                    eval_sum += 1                                       # eval_sum为测试数据的数目
                    eval_list.append(name_path + "\t%d" % class_label + "\n")
                else:
                    trainer_sum += 1 
                    trainer_list.append(name_path + "\t%d" % class_label + "\n")#trainer_sum测试数据的数目
                class_sum += 1                                          #每类图片的数目
                all_class_images += 1                                   #所有类图片的数目
            
            # 说明的json文件的class_detail数据
            class_detail_list['class_name'] = class_dir             #类别名称
            class_detail_list['class_label'] = class_label          #类别标签
            class_detail_list['class_eval_images'] = eval_sum       #该类数据的测试集数目
            class_detail_list['class_trainer_images'] = trainer_sum #该类数据的训练集数目
            class_detail.append(class_detail_list)  
            #初始化标签列表
            train_parameters['label_dict'][str(class_label)] = class_dir
            class_label += 1
            
    #初始化分类数
    train_parameters['class_dim'] = class_dim
    print(train_parameters)
    #乱序  
    random.shuffle(eval_list)
    with open(eval_list_path, 'a') as f:
        for eval_image in eval_list:
            f.write(eval_image) 
    #乱序        
    random.shuffle(trainer_list) 
    with open(train_list_path, 'a') as f2:
        for train_image in trainer_list:
            f2.write(train_image) 

    # 说明的json文件信息
    readjson = {}
    readjson['all_class_name'] = data_list_path                  #文件父目录
    readjson['all_class_images'] = all_class_images
    readjson['class_detail'] = class_detail
    jsons = json.dumps(readjson, sort_keys=True, indent=4, separators=(',', ': '))
    with open(train_parameters['readme_path'],'w') as f:
        f.write(jsons)
    print ('生成数据列表完成！')

生成数据列表

'''
参数初始化
'''
src_path=train_parameters['src_path']
target_path=train_parameters['target_path']
train_list_path=train_parameters['train_list_path']
eval_list_path=train_parameters['eval_list_path']
batch_size=train_parameters['train_batch_size']
'''
解压原始数据到指定路径
'''
unzip_data(src_path,target_path)

'''
划分训练集与验证集，乱序，生成数据列表
'''
#每次生成数据列表前，首先清空train.txt和eval.txt
with open(train_list_path, 'w') as f: 
    f.seek(0)
    f.truncate() 
with open(eval_list_path, 'w') as f: 
    f.seek(0)
    f.truncate() 
    
#生成数据列表   
get_data_list(target_path,train_list_path,eval_list_path)

输出

{'input_size': [3, 224, 224], 'class_dim': 25, 'src_path': 'data/data55032/archive_train.zip', 'target_path': '/home/aistudio/data/dataset', 'train_list_path': './train.txt', 'eval_list_path': './eval.txt', 'label_dict': {'0': 'Cats Eye', '1': 'Diamond', '2': 'Labradorite', '3': 'Quartz Beer', '4': 'Variscite', '5': 'Almandine', '6': 'Fluorite', '7': 'Sapphire Blue', '8': 'Jade', '9': 'Danburite', '10': 'Zircon', '11': 'Pearl', '12': 'Kunzite', '13': 'Malachite', '14': 'Tanzanite', '15': 'Emerald', '16': 'Benitoite', '17': 'Carnelian', '18': 'Hessonite', '19': 'Iolite', '20': 'Onyx Black', '21': 'Alexandrite', '22': 'Beryl Golden', '23': 'Rhodochrosite', '24': 'Garnet Red'}, 'readme_path': '/home/aistudio/data/readme.json', 'num_epochs': 1, 'train_batch_size': 16, 'learning_strategy': {'lr': 0.001}}
生成数据列表完成！

封装数据

class Reader(Dataset):
    def __init__(self, data_path, mode='train'):
        """
        数据读取器
        param data_path: 数据集所在路径
        param mode: train or eval
        """
        super().__init__()
        self.data_path = data_path
        self.img_paths = []
        self.labels = []

        if mode == 'train':
            with open(os.path.join(self.data_path, "train.txt"), "r", encoding="utf-8") as f:
                self.info = f.readlines()
            for img_info in self.info:
                img_path, label = img_info.strip().split('\t')
                self.img_paths.append(img_path)
                self.labels.append(int(label))

        else:
            with open(os.path.join(self.data_path, "eval.txt"), "r", encoding="utf-8") as f:
                self.info = f.readlines()
            for img_info in self.info:
                img_path, label = img_info.strip().split('\t')
                self.img_paths.append(img_path)
                self.labels.append(int(label))


    def __getitem__(self, index):
        """
        获取一组数据 :param index: 文件索引号
        :return:
        """
        # 第一步打开图像文件并获取label值
        img_path = self.img_paths[index]
        img = Image.open(img_path)
        if img.mode != 'RGB':
            img = img.convert('RGB') 
        img = img.resize((224, 224), Image.BILINEAR)
        img = np.array(img).astype('float32')
        img = img.transpose((2, 0, 1)) / 255
        label = self.labels[index]
        label = np.array([label], dtype="int64")
        return img, label

    def print_sample(self, index: int = 0):
        print("文件名", self.img_paths[index], "\t标签值", self.labels[index])

    def __len__(self):
        """
        返回数据集样本个数
        """
        return len(self.img_paths)

训练&测试数据加载

train_dataset = Reader('/home/aistudio',mode='train')

eval_dataset = Reader('/home/aistudio/',mode='eval')

#训练数据加载
train_loader = paddle.io.DataLoader(train_dataset, batch_size=16, shuffle=True)
#测试数据加载
eval_loader = paddle.io.DataLoader(eval_dataset, batch_size = 8, shuffle=False)

train_dataset.print_sample(200)
print(train_dataset.__len__())
eval_dataset.print_sample(0)
print(eval_dataset.__len__())
print(eval_dataset.__getitem__(10)[0].shape)
print(eval_dataset.__getitem__(10)[1].shape)

Batch=0
Batchs=[]
all_train_accs=[]
def draw_train_acc(Batchs, train_accs):
    title="training accs"
    plt.title(title, fontsize=24)
    plt.xlabel("batch", fontsize=14)
    plt.ylabel("acc", fontsize=14)
    plt.plot(Batchs, train_accs, color='green', label='training accs')
    plt.legend()
    plt.grid()
    plt.show()

all_train_loss=[]
def draw_train_loss(Batchs, train_loss):
    title="training loss"
    plt.title(title, fontsize=24)
    plt.xlabel("batch", fontsize=14)
    plt.ylabel("loss", fontsize=14)
    plt.plot(Batchs, train_loss, color='red', label='training loss')
    plt.legend()
    plt.grid()
    plt.show()

定义模型

定义网络

继承paddle.nn.Layer

#定义DNN网络
class MyDNN(paddle.nn.Layer):
    def __init__(self):
        super(MyDNN,self).__init__()
        self.linear1 = paddle.nn.Linear(in_features=3*224*224, out_features=1024)
        self.relu1 = paddle.nn.ReLU()

        self.linear2 = paddle.nn.Linear(in_features=1024, out_features=512)
        self.relu2 = paddle.nn.ReLU()

        self.linear3 = paddle.nn.Linear(in_features=512, out_features=128)
        self.relu3 = paddle.nn.ReLU()

        self.linear4 = paddle.nn.Linear(in_features=128, out_features=25)

    def forward(self,input):        # forward 定义执行实际运行时网络的执行逻辑
        # input.shape (16, 3, 224, 224)
        x = paddle.reshape(input, shape=[-1,3*224*224]) #-1 表示这个维度的值是从x的元素总数和剩余维度推断出来的，有且只能有一个维度设置为-1
        # print(x.shape)
        x = self.linear1(x)
        x = self.relu1(x)
        # print('1', x.shape)
        x = self.linear2(x)
        x = self.relu2(x)
        # print('2',x.shape)
        x = self.linear3(x)
        x = self.relu3(x)
        # print('3',x.shape)
        y = self.linear4(x)
        # print('4',y.shape)
        return y

训练模型

model=MyDNN() #模型实例化
model.train() #训练模式
cross_entropy = paddle.nn.CrossEntropyLoss()#定义损失
opt=paddle.optimizer.SGD(learning_rate=0.001, parameters=model.parameters())#paddle优化器

epochs_num=train_parameters['num_epochs'] #迭代次数
for pass_num in range(train_parameters['num_epochs']):
    for batch_id,data in enumerate(train_loader()):
        image = data[0]
        label = data[1]

        predict=model(image) #数据传入model

        loss=cross_entropy(predict,label)#计算损失
        acc=paddle.metric.accuracy(predict,label)#计算精度
        
        if batch_id!=0 and batch_id%5==0:#绘图
            Batch = Batch+5 
            Batchs.append(Batch)
            all_train_loss.append(loss.numpy()[0])
            all_train_accs.append(acc.numpy()[0])
            print("train_pass:{},batch_id:{},train_loss:{},train_acc:{}".format(pass_num,batch_id,loss.numpy(),acc.numpy()))
        
        loss.backward()       #反向传播
        opt.step()	
        opt.clear_grad()   #opt.clear_grad()来重置梯度

paddle.save(model.state_dict(),'MyDNN')#保存模型

draw_train_acc(Batchs,all_train_accs)
draw_train_loss(Batchs,all_train_loss)

第一轮输出

train_pass:0,batch_id:5,train_loss:[3.271103],train_acc:[0.]
train_pass:0,batch_id:10,train_loss:[3.4631658],train_acc:[0.]
train_pass:0,batch_id:15,train_loss:[3.3294864],train_acc:[0.125]
train_pass:0,batch_id:20,train_loss:[3.1694913],train_acc:[0.0625]
train_pass:0,batch_id:25,train_loss:[2.9359517],train_acc:[0.25]
train_pass:0,batch_id:30,train_loss:[3.2490888],train_acc:[0.]
train_pass:0,batch_id:35,train_loss:[3.0921557],train_acc:[0.125]
train_pass:0,batch_id:40,train_loss:[3.1657226],train_acc:[0.125]
train_pass:0,batch_id:45,train_loss:[3.0306082],train_acc:[0.0625]

第一轮的图

泛化性能评估

用测试集

#模型评估
para_state_dict = paddle.load("MyDNN")
model = MyDNN()
model.set_state_dict(para_state_dict) #加载模型参数
model.eval() #验证模式

accs = []

for batch_id,data in enumerate(eval_loader()):#测试集
    image=data[0]
    label=data[1]     
    predict=model(image)       
    acc=paddle.metric.accuracy(predict,label)
    accs.append(acc.numpy()[0])
    avg_acc = np.mean(accs)
print("当前模型在验证集上的准确率为:",avg_acc)

应用模型预测

import os
import zipfile

def unzip_infer_data(src_path,target_path):
    '''
    解压预测数据集
    '''
    if(not os.path.isdir(target_path)):     
        z = zipfile.ZipFile(src_path, 'r')
        z.extractall(path=target_path)
        z.close()


def load_image(img_path):
    '''
    预测图片预处理
    '''
    img = Image.open(img_path) 
    if img.mode != 'RGB': 
        img = img.convert('RGB') 
    img = img.resize((224, 224), Image.BILINEAR)
    img = np.array(img).astype('float32') 
    img = img.transpose((2, 0, 1))  # HWC to CHW 
    img = img/255                # 像素值归一化 
    return img


infer_src_path = '/home/aistudio/data/data55032/archive_test.zip'
infer_dst_path = '/home/aistudio/data/archive_test'
unzip_infer_data(infer_src_path,infer_dst_path)

'''
模型预测
'''
para_state_dict = paddle.load("MyDNN")
model = MyDNN()
model.set_state_dict(para_state_dict) #加载模型参数
model.eval() #训练模式

#展示预测图片
infer_path='data/archive_test/alexandrite_3.jpg'
img = Image.open(infer_path)
plt.imshow(img)          #根据数组绘制图像
plt.show()               #显示图像

#对预测图片进行预处理
infer_imgs = []
infer_imgs.append(load_image(infer_path))
infer_imgs = np.array(infer_imgs)

label_dic = train_parameters['label_dict']

for i in range(len(infer_imgs)):
    data = infer_imgs[i]
    dy_x_data = np.array(data).astype('float32')
    dy_x_data=dy_x_data[np.newaxis,:, : ,:]
    img = paddle.to_tensor (dy_x_data)
    out = model(img)
    lab = np.argmax(out.numpy())  #argmax():返回最大数的索引

    print("第{}个样本,被预测为：{},真实标签为：{}".format(i+1,label_dic[str(lab)],infer_path.split('/')[-1].split("_")[0]))
        
print("结束")