资源简介
用于指静脉识别,Python语言,keras架构,CNN卷积神经网络,识别率可以达到97%。禁止转载。
代码片段和文件信息
#-*- coding:utf-8 -*-
import keras
from keras.preprocessing.image import ImageDataGenerator
from keras.models import Sequential
from keras.layers import Conv2D MaxPooling2D
from keras.layers import Dropout Flatten Dense
from keras.utils import plot_model
import matplotlib.pyplot as plt
from keras.utils import multi_gpu_model
import os
from read_data import get_data
#from get_tfr_data import get_data
from sklearn.cross_validation import cross_val_score # K折交叉验证模块
‘消除警告‘
os.environ[‘TF_CPP_MIN_LOG_LEVEL‘]=‘2‘
size = 128 #图片尺寸
def Model():
# 搭建卷积神经网络
model = Sequential()
model.add(Conv2D(filters = 8
kernel_size = (7 7)
strides = (22)
padding = ‘same‘
input_shape = (sizesize1)
activation = ‘relu‘ ))
model.add(MaxPooling2D(pool_size=(3 3)strides=(11)))
model.add(Dropout(0.2))
# 重复构造,搭建深度网络
model.add(Conv2D(16 kernel_size = (5 5) strides = (11) padding = ‘same‘ activation = ‘relu‘))
model.add(MaxPooling2D(pool_size = (2 2)strides=(11)))
model.add(Dropout(0.2))
‘‘‘
model.add(Conv2D(filters = 16
kernel_size = (2 2)
strides = (11)
padding = ‘same‘
input_shape = (sizesize1)
activation = ‘relu‘ ))
model.add(MaxPooling2D(pool_size=(2 2)strides=(11)))
#model.add(Dropout(0.5))
# 重复构造,搭建深度网络
model.add(Conv2D(16
kernel_size = (2 2)
strides = (11)
padding = ‘same‘
activation = ‘relu‘))
model.add(MaxPooling2D(pool_size = (2 2)strides=(11)))
#model.add(Dropout(0.5))
model.add(Conv2D(8
kernel_size = (2 2)
strides = (11)
padding = ‘same‘
activation = ‘relu‘))
#model.add(Conv2D(64 kernel_size = (3 3) strides = (11) padding = ‘same‘ activation = ‘relu‘))
#model.add(Conv2D(32 kernel_size = (3 3) strides = (11) padding = ‘same‘ activation = ‘relu‘))
#model.add(MaxPooling2D(pool_size = (3 3)strides=(22)))
model.add(Dropout(0.5))
‘‘‘
# 把当前节点展平
model.add(Flatten())
#
相关资源
- LSTM股价预测(python).zip
- keras库对Kaggle竞赛猫狗图像分类卷积神
- GRU神经网络 Python代码
- python3使用tensorflow构建CNN卷积神经网络
- keras .whl文件 用于python3
- CNN卷积神经网络python代码
- 基于Keras的股票价格预测
- 神经网络-二分类问题(IMDB) Keras
- Attention-CNN(Jianlong-Fu 大神制作)
- 基于卷积神经网络的手写数字识别
- Tensorflow之CNN实现CIFAR-10图像的分类p
- cython_bbox.so
- 适合的新手-CNN代码
- keras_inception_v4_finetune.py
- densenet121.py
- mnist_acgan.py
- CNN网络识别Mnist的源码,有详细注释,
- CNN_源代码
- faster rcnn end-to-end loss曲线的绘制
- 卷积神经网络回归模型
- Python Keras库 安装包
- Python-Keras实现实时语义分割的深层神
- Python+Tensorflow+CNN实现车牌识别的
- 利用keras实现的cnn卷积神经网络对手写
- AI智能五子棋Python代码
- keras上LSTM长短期记忆网络金融时序预
- CNN卷积神经网络TensorFlow代码
- 深度学习 莫烦 Keras源代码
- DeepLab-ResNet-101
- Python-Keras实现的DeepSpeech端到端语音识
评论
共有 条评论