首页 > 专利学习

fer2013数据集_Tensorflow2.0CNN实战（4）-Fer2013数据集

fer2013数据集_Tensorflow2.0CNN实战（4）-Fer2013数据集上篇讲到了，如何在kaggle上进⾏GPU加速，这边就使⽤免费的GPU，尝试⼀波。所⽤数据集是fer2013，本篇作为⼀个引⼦来讲，下⼀篇会对FER2013 PLUS数据集进⾏训练，准确度⽐fer2013增加了很多。

1、数据集介绍

fer2013数据集由35886张⼈脸不同表情图⽚组成，其中训练集28708张，验证集和测试集各3589张。每张图⽚的⼤⼩是48*48像素⼤⼩，表情分为7种：0 anger ⽣⽓； 1 disgust 厌恶； 2 fear 恐惧； 3 happy 开⼼；4normal； 5 sad 伤⼼；6 surprised 惊讶。⽹上下载的数据集是⼀个csv⽂件，需要对其进⾏处理，这⾥将每⼀种表情按照不⽤⽂件存储。

2、分⽂件保存数据集

打⼊库函数

import matplotlib as mpl

import matplotlib.pyplot as plt

%matplotlib inline

import numpy as np

import pandas as pd

import os

import sys

import time

import tensorflow as tf

from tensorflow import keras

import scipy.misc as sm

import scipy

from PIL import Image

按照训练集、验证集、测试集分成不同的csv⽂件

import csv

database_path = 'D:/pythonstudy/jupyter-notebook/face/fer2013_new/'

datasets_path = './fer2013_new/'

csv_file = database_path+'fer2013_newnew.csv'

train_csv = datasets_path+'train.csv'

val_csv = datasets_path+'val.csv'

test_csv = datasets_path+ 'test.csv'

with open(csv_file) as f:

csvr = ader(f)

header = next(csvr)

汽车喷水电机print(header)

rows = [row for row in csvr]

trn = [row[:-1] for row in rows if row[-1] == 'Training']

csv.writer(open(train_csv, 'w+'), lineterminator='n').writerows([header[:-1]] + trn)

print(len(trn))

val = [row[:-1] for row in rows if row[-1] == 'PublicTest']

csv.writer(open(val_csv, 'w+'), lineterminator='n').writerows([header[:-1]] + val)

print(len(val))

tst = [row[:-1] for row in rows if row[-1] == 'PrivateTest']

csv.writer(open(test_csv, 'w+'), lineterminator='n').writerows([header[:-1]] + tst)

print(len(tst))

对不同的csv⽂件进⾏不⽤表情⽂件保存

datasets_path = r'.fer2013_new'

train_csv = os.path.join(datasets_path, 'train.csv')

val_csv = os.path.join(datasets_path, 'val.csv')

test_csv = os.path.join(datasets_path, 'test.csv')

train_set = os.path.join(datasets_path, 'train')

val_set = os.path.join(datasets_path, 'val')

test_set = os.path.join(datasets_path, 'test')

for save_path, csv_file in [(train_set, train_csv), (val_set, val_csv), (test_set, test_csv)]: if not ists(save_path):

os.makedirs(save_path)

num = 1

with open(csv_file) as f:

csvr = ader(f)

header = next(csvr)

for i, (label, pixel) in enumerate(csvr):

pixel = np.asarray([float(p) for p in pixel.split()]).reshape(48, 48)

subfolder = os.path.join(save_path, label)

if not ists(subfolder):

os.makedirs(subfolder)

空气源热泵热水系统

im = Image.fromarray(pixel).convert('L')

image_name = os.path.join(subfolder, '{:05d}.jpg'.format(i))

#print(image_name)

im.save(image_name)

高压直流供电保存后的⽂件见下图

3、kaggle上传数据并训练

将训练集、验证集、测试集进⾏压缩并上传到kaggle上，具体如何上传可参考上篇‘kaggle免费GPU加速’，并⽤gpu加速进⾏训练。

3.1按照顺序定义class_names

class_names=['anger','disgust','fear','happy',

'normal','sad','surprised',]

train_folder='../input/fer2013/train'

val_folder='../input/fer2013/val'

test_folder='../input/fer2013/test'

3.2增强数据集

height=48

width=48

channels=1

batch_size=64

num_classes=7

train_datagen=keras.preprocessing.image.ImageDataGenerator( rescale=1./255,

rotation_range=40,

width_shift_range=0.2,

height_shift_range=0.2,

shear_range=0.2,

zoom_range=0.2,

horizontal_flip=True,

fill_mode='nearest')

train_generator=train_datagen.flow_from_directory(

train_folder,

target_size=(height,width),

batch_size=batch_size,

蜂窝纸芯

seed=7,

音调控制电路shuffle=True,

class_mode='categorical')

valid_datagen=keras.preprocessing.image.ImageDataGenerator( rescale=1./255)

valid_generator=train_datagen.flow_from_directory(

val_folder,

target_size=(height,width),

batch_size=batch_size,

seed=7,

shuffle=False,

class_mode='categorical')

test_datagen=keras.preprocessing.image.ImageDataGenerator( rescale=1./255)

相册加工设备

test_generator=test_datagen.flow_from_directory(

test_folder,

target_size=(height,width),

batch_size=batch_size,

seed=7,

shuffle=False,

class_mode='categorical')

train_num=train_generator.samples

valid_num=valid_generator.samples

print(train_num,valid_num)

3.3、建⽴模型结构

model = dels.Sequential()

model.add(keras.layers.Conv2D(filters=32,kernel_size=3,

padding='same',activation='relu',

input_shape=(width,height,3)))