juypter 基于keras训练的笑脸检测+口罩人脸识别

最新推荐文章于 2022-01-24 17:33:13 发布

chears

最新推荐文章于 2022-01-24 17:33:13 发布

阅读量1k

点赞数

分类专栏：人工智能文章标签：神经网络

本文链接：https://blog.csdn.net/weixin_43749019/article/details/107283716

版权

这篇博客详细介绍了如何使用Keras训练基于genki4k数据集的笑脸检测模型，包括数据预处理、构建小型CNN、数据增强等步骤，并讨论了防止过拟合的方法。接着，博主展示了如何转换数据集进行口罩人脸识别，提供了相应的训练流程。

摘要由CSDN通过智能技术生成

一、笑脸数据集（genki4k）正负样本的划分、模型训练和测试的过程（至少包括SVM、CNN），输出模型训练精度和测试精度（F1-score和ROC）

1.训练数据

1) 导入Keras

import keras
keras.__version__

'2.3.1'

2) 读取数据集，训练

import os, shutil
# The path to the directory where the original
# dataset was uncompressed
original_dataset_dir = 'smileface/genki4k'

# The directory where we will
# store our smaller dataset
base_dir = 'smileface/1'
os.mkdir(base_dir)

# Directories for our training,
# validation and test splits
train_dir = os.path.join(base_dir, 'train')
os.mkdir(train_dir)
validation_dir = os.path.join(base_dir, 'validation')
os.mkdir(validation_dir)
test_dir = os.path.join(base_dir, 'test')
os.mkdir(test_dir)

# Directory with our training smile pictures
train_smile_dir = os.path.join(train_dir, 'smiles')
os.mkdir(train_smile_dir)

# Directory with our training nosmile pictures
train_nosmile_dir = os.path.join(train_dir, 'no_smiles')
os.mkdir(train_nosmile_dir)

# Directory with our validation smile pictures
validation_smile_dir = os.path.join(validation_dir, 'smiles')
os.mkdir(validation_smile_dir)

# Directory with our validation nosmile pictures
validation_nosmile_dir = os.path.join(validation_dir, 'no_smiles')
os.mkdir(validation_nosmile_dir)

# Directory with our validation smile pictures
test_smile_dir = os.path.join(test_dir, 'smiles')
os.mkdir(test_smile_dir)

# Directory with our validation nosmile pictures
test_nosmile_dir = os.path.join(test_dir, 'no_smiles')
os.mkdir(test_nosmile_dir)

3）将笑脸数据存入对应文件夹

在这里插入图片描述

print('total training smile images:', len(os.listdir(train_smile_dir)))
print('total training nosmile images:', len(os.listdir(train_nosmile_dir)))
print('total validation smile images:', len(os.listdir(validation_smile_dir)))
print('total validation nosmile images:', len(os.listdir(validation_nosmile_dir)))
print('total test smile images:', len(os.listdir(test_smile_dir)))
print('total test nosmile images:', len(os.listdir(test_nosmile_dir)))

total training smile images: 2162
total training nosmile images: 1837
total validation smile images: 2162
total validation nosmile images: 1837
total test smile images: 2162
total test nosmile images: 1837

4) 构建小型卷积神经网络

from keras import layers
from keras import models

model = models.Sequential()
model.add(layers.Conv2D(32, (3, 3), activation='relu',
                        input_shape=(150, 150, 3)))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(128, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Flatten())
model.add(layers.Dense(512, activation='relu'))
model.add(layers.Dense(1, activation='sigmoid'))

了解征图的尺寸是如何随着每一层变化的

model.summary()

Model: "sequential_2"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
===============================================================