您现在的位置是:热点 >>正文
教你使用TensorFlow2对阿拉伯语手写字符数据集进行识别
热点53581人已围观
简介...
在本教程中,我们将使用 TensorFlow (Keras API) 实现一个用于多分类任务的深度学习模型,该任务需要对阿拉伯语手写字符数据集进行识别。
数据集下载地址:https://www.kaggle.com/mloey1/ahcd1
数据集介绍
该数据集由 60 名参与者书写的16,800 个字符组成,年龄范围在 19 至 40 岁之间,90% 的参与者是右手。
每个参与者在两种形式上写下每个字符(从“alef”到“yeh”)十次,如图 7(a)和 7(b)所示。表格以 300 dpi 的分辨率扫描。使用 Matlab 2016a 自动分割每个块以确定每个块的坐标。该数据库分为两组:训练集(每类 13,440 个字符到 480 个图像)和测试集(每类 3,360 个字符到 120 个图像)。数据标签为1到28个类别。在这里,所有数据集都是CSV文件,表示图像像素值及其相应标签,并没有提供对应的图片数据。
导入模块
import numpy as np import pandas as pd #允许对dataframe使用display() from IPython.display import display # 导入读取和处理图像所需的库 import csv from PIL import Image from scipy.ndimage import rotate
读取数据
# 训练数据images letters_training_images_file_path = "../input/ahcd1/csvTrainImages 13440x1024.csv" # 训练数据labels letters_training_labels_file_path = "../input/ahcd1/csvTrainLabel 13440x1.csv" # 测试数据images和labels letters_testing_images_file_path = "../input/ahcd1/csvTestImages 3360x1024.csv" letters_testing_labels_file_path = "../input/ahcd1/csvTestLabel 3360x1.csv" # 加载数据 training_letters_images = pd.read_csv(letters_training_images_file_path, header=None) training_letters_labels = pd.read_csv(letters_training_教你labels_file_path, header=None) testing_letters_images = pd.read_csv(letters_testing_images_file_path, header=None) testing_letters_labels = pd.read_csv(letters_testing_labels_file_path, header=None) print("%d个32x32像素的训练阿拉伯字母图像。" %training_letters_images.shape[0]) print("%d个32x32像素的测试阿拉伯字母图像。" %testing_letters_images.shape[0]) training_letters_images.head()
13440个32x32像素的训练阿拉伯字母图像。3360个32x32像素的测试阿拉伯字母图像。
查看训练数据的head
np.unique(training_letters_labels) array([ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17,18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28], dtype=int32)
下面需要将csv值转换为图像,我们希望展示对应图像的像素值图像。
def convert_values_to_image(image_values, display=False): image_array = np.asarray(image_values) image_array = image_array.reshape(32,32).astype('uint8') # 原始数据集被反射,因此我们将使用np.flip翻转它,然后通过rotate旋转,以获得更好的图像。 image_array = np.flip(image_array, 0) image_array = rotate(image_ar使用ray, -90) new_image = Image.fromarray(image_array) if display == True: new_image.show() return new_image convert_values_to_image(training_letters_images.loc[0], True)
这是一个字母f。
下面,我们将进行数据预处理,主要进行图像标准化,我们通过将图像中的每个像素除以255来重新缩放图像,标准化到[0,1]
training_letters_images_scaled = training_letters_images.values.astype('float32')/255 training_letters_labels = training_letters_labels.values.astype('int32') testing_letters_images_scaled = testing_letters_images.values.astype('float32')/255 testing_letters_labels = testing_letters_labels.values.astype('int32') print("Training images of letters after scaling") print(training_letters_images_scaled.shape) training_letters_images_scaled[0:5]
输出如下
Training images of letters after scaling (13440, 1024)
从标签csv文件我们可以看到,这是一个多类分类问题。下一步需要进行分类标签编码,建议将类别向量转换为矩阵类型。
输出形式如下:将1到28,变成0到27类别。从“alef”到“yeh”的字母有0到27的分类号。to_categorical就是将类别向量转换为二进制(只有0和1)的矩阵类型表示
在这里,我们将使用keras的一个热编码对这些类别值进行编码。
一个热编码将整数转换为二进制矩阵,其中数组仅包含一个“1”,其余元素为“0”。
from keras.utils import to_categorical # one Tehot encoding number_of_classes = 28 training_letters_labels_encoded = to_categorical(training_letters_labels-1, num_classes=number_of_classes) testing_letters_labels_encoded = to_categorical(testing_letters_labels-1, num_classes=number_of_classes)
# (13440, 1024)
下面将输入图像重塑为32x32x1,因为当使用TensorFlow作为后端时,Keras CNN需要一个4D数组作为输入,并带有形状(nb_samples、行、列、通道)
其中 nb_samples对应于图像(或样本)的总数,而行、列和通道分别对应于每个图像的行、列和通道的数量。
# reshape input letter images to 32x32x1 training_letters_images_scaled = training_letters_images_scaled.reshape([-1, 32, 32, 1]) testing_letters_images_scaled = testing_letters_images_scaled.reshape([-1, 32, 32, 1]) print(training_letters_images_scaled.shape, training_letters_labels_encoded.shape, testing_letters_images_scaled.shape, testing_letters_labels_encoded.shape) # (13440, 32, 32, 1) (13440, 28) (3360, 32, 32, 1) (3360, 28)
因此,我们将把输入图像重塑成4D张量形状(nb_samples,32,32,1),因为我们图像是32x32像素的灰度图像。
#将输入字母图像重塑为32x32x1 training_letters_images_scaled = training_letters_images_scaled.reshape([-1, 32, 32, 1]) testing_letters_images_scaled = tesnsting_letters_images_scaled.reshape([-1, 32, 32, 1]) print(training_letters_images_scaled.shape, training_letters_labels_encoded.shape, testing_letters_images_scaled.shape, testing_letters_labels_encoded.shape)
设计模型结构
from keras.models import Sequential from keras.layers import Conv2D, MaxPooling2D, GlobalAveragePooling2D, BatchNormalization, Dropout, Dense def create_model(optimizer='adam', kernel_initializer='he_normal', activation='relu'): # create model model = Sequential() model.add(Conv2D(filters=16, kernel_size=3, padding='same', input_shape=(32, 32, 1), kernel_initializer=kernel_initializer, activation=activation)) model.add(BatchNormalization()) model.add(MaxPooling2D(pool_size=2)) model.add(Dropout(0.2)) model.add(Conv2D(filters=32, kernel_size=3, padding='same', kernel_initializer=kernel_initializer, activation=activation)) model.add(BatchNormalization()) model.add(MaxPooorling2D(pool_size=2)) model.add(Dropout(0.2)) model.add(Conv2D(filters=64, kernel_size=3, padding='same', kernel_initializer=kernel_initializer, activation=activation)) model.add(BatchNormalization()) model.add(MaxPooling2D(pool_size=2)) model.add(Dropout(0.2)) model.add(Conv2D(filters=128, kernel_size=3, padding='same', kernel_initializer=kernel_initializer, activation=activation)) model.add(BatchNormalization()) model.add(MaxPooling2D(pool_size=2)) model.add(Dropout(0.2)) model.add(GlobalAveragePooling2D()) #Fully connected final layer model.add(Dense(28, activation='softmax')) # Compile model model.compile(loss='categorical_crossentropy', metrics=['accuracy'], optimizer=optimizer) return model 「模型结构」 第一隐藏层是卷积层。该层有16个特征图,大小为3×3和一个激活函数,它是relu。这是输入层,需要具有上述结构的图像。 第二层是批量标准化层,它解决了特征分布在训练和测试数据中的变化,BN层添加在激活函数前,对输入激Fl活函数的输入进行归一化。这样解决了输入数据发生偏移和增大的影响。 第三层是MaxPooling层。最大池层用于对输入进行下采样,使模型能够对特征进行假设,从而减少过拟合。它还减少了参数的学习次数,减少了训练时间。 下一层是使用dropout的正则化层。它被配置为随机排除层中20%的神经元,以减少过度拟合。 另一个隐藏层包含32个要素,大小为3×3和relu激活功能,从图像中捕捉更多特征。 其他隐藏层包含64和128个要素,大小为3×3和一个relu激活功能, 重复三次卷积层、MaxPooling、批处理规范化、正则化和* GlobalAveragePooling2D层。 最后一层是具有(输出类数)的输出层,它使用softmax激活函数,因为我们有多个类。每个神经元将给出该类的概率。 使用分类交叉熵作为损失函数,因为它是一个多类分类问题。使用精确度作为衡量标准来提高神经网络的性能。 model = create_model(optimizer='Adam', kernel_initializer='uniform', activation='relu') model.summary()
「Keras支持在Keras.utils.vis_utils模块中绘制模型,该模块提供了使用graphviz绘制Keras模型的实用函数」
import pydot from keras.utils import plot_model plot_model(model, to_file="model.png", show_shapes=True) from IPython.display import Image as IPythonImage display(IPythonImage('model.png'))
训练模型,使用batch_size=20来训练模型,对模型进行15个epochs阶段的训练。
from keras.callbacks import ModelCheckpoint # 使用检查点来保存稍后使用的模型权重。 checkpointer = ModelCheckpoint(filepath='weights.hdf5', verbose=1, save_best_only=True) history = model.fit(training_letters_images_scaled, training_letters_labels_encoded,validation_data=(testing_letters_images_scaled,testing_letters_labels_encoded),epochs=15, batch_size=20, ow;verbose=1, callbacks=[checkpointer])
训练结果如下所示:
最后Epochs绘制损耗和精度曲线。
import matplotlib.pyplot as plt def plot_loss_accuracy(history): # Loss plt.figure(figsize=[8,6]) plt.plot(history.history['loss'],'r',linewidth=3.0) plt.plot(history.history['val_loss'],'b',linewidth=3.0) plt.legend(['Training loss', 'Validation Loss'],fontsize=18) plt.xlabel('Epochs ',fontsize=16) plt.ylabel('Loss',fontsize=16) plt.title('Loss Curves',fontsize=16) # Accuracy plt.figure(figsize=[8,6]) plt.plot(history.history['accuracy'],'r',linewidth=3.0) plt.plot(history.history['val_accuracy'],'b',linewidth=3.0) plt.legend(['Training Accuracy', 'Validation Accuracy'],fontsize=18) plt.xlabel('Epochs ',fontsize=16) plt.ylabel('Accuracy',fontsize=16) plt.title('Accuracy Curves',fontsize=16) &nb2对sp;plot_loss_accuracy(history)
「加载具有最佳验证损失的模型」 # 加载具有最佳验证损失的模型 model.load_weights('weights.hdf5') metrics = model.evaluate(testing_letters_images_scaled, testing_letters_labels_encoded, verbose=1) print("Test Accuracy: {}".format(metrics[1])) print("Test Loss: {}".format(metrics[0]))
输出如下:
3360/3360 [==============================] - 0s 87us/step Test Accuracy: 0.9678571224212646 Test Loss: 0.11759862171020359
打印混淆矩阵。
from sklearn.metrics import classification_report def get_predicted_classes(model, data, labels=None): image_predictions = model.predict(data) predicted_classes = np.argmax(image_predictions, axis=1) true_classes = np.argmax(labels, axis=1) return predicted_classes, true_classes, image_predictions def get_classification_report(y_true, y_pred): print(classification_report(y_true, y_pred)) y_pred, y_true, image_predictions = get_predicted_classes(model, testing_letters_images_scaled, 阿拉testing_letters_labels_encoded) get_classification_report(y_true, y_pred)
输出如下:
precision recall f1-score support 0 1.00 0.98 0.99 120 1 1.00 0.98 0.99 120 2 0.80 0.98 0.88 120 3 0.98 0.88 0.93 120 4 0.99 0.97 0.98 &nbs伯语p; 120 5 0.92 0.99 0.96 120 6 0.94 0.97 0.95 120 7 0.94 0.95 0.95 120 8 0.96 0.88 0.92 120 9 0.90 1.00 0.94 120 10 0.94 0.90 0.92 120&手写nbsp; 11 0.98 1.00 0.99 120 12 0.99 0.98 0.99 120 13 0.96 0.97 0.97 120 14 1.00 0.93 0.97 120 15 0.94 0.99 0.97 120 16 1.00 0.93 0.96 120  字符; 17 0.97 0.97 0.97 120 18 1.00 0.93 0.96 120 19 0.92 0.95 0.93 120 20 0.97 0.93 0.94 120 21 0.99 0.96 0.97 120 22 0.99 0.98 0.99 120 23 &数据nbsp; 0.98 0.99 0.99 120 24 0.95 0.88 0.91 120 25 0.94 0.98 0.96 120 26 0.95 0.97 0.96 120 27 0.98 0.99 0.99 120 accuracy 0.96 3360 macro avg 0.96 0.96&nb集进sp; 0.96 3360 ghted avg 0.96 0.96 0.96 3360
最后绘制随机几个相关预测的图片
indices = np.random.randint(0, testing_letters_labels.shape[0], size=49) y_pred = np.argmax(model.predict(training_letters_images_scaled), axis=1) for i, idx in enumerate(indices): plt.subplot(7,7,i+1) image_array = training_letters_images_scaled[idx][:,:,0] image_array = np.flip(image_array, 0) image_array = rotate(image_array, -90) plt.imshow(image_array, cmap='gray') plt.title("Pred: {} - Label: {}".format(y_pred[idx], (training_letters_labels[idx] -1))) plt.xticks([]) plt.yticks([]) plt.show()
Tags:
相关文章
标签: 苹果 iPhone 17 M5芯片 Apple Intelligence
热点云梦级1月9日云梦级消息,苹果公司今日发布 SEC 文件,官宣将于美东时间 2 月 24 日举行年度股东大会。文件中,苹果 CEO 蒂姆・库克发布了一封致股东感谢信。 内容如下: 致股东: 对苹果来说...
【热点】
阅读更多数据分析的12个神话被揭穿!
热点从数据问题到人员需求再到技术组合,数据分析的误解比比皆是。下面我们来看看如何利用数据科学来实现真正的业务成果。 在IT领域,炒作越大,误解越多,数据分析也不例外。分析是当今信息技术最热门的方面之一,可...
【热点】
阅读更多标签: 新氧 SY 美容治疗服务 SoYoung
热点云梦级6月3日云梦级消息,新氧 (Nasdaq: SY) 日前发布截至2026年3月31日的财报。财报显示,新氧2026年第一季度营收为4.33亿元(约6270万美元),较上年同期的3亿元增长46%。...
【热点】
阅读更多
热门文章
最新文章
友情链接
- 标签: 康师傅
- https://www.safetv.go.kr/base/search/view?searchWord=谷歌快排操作流程【tg:seogoogle987】6aX
- https://www.yidian51.com/taobao/谷歌快排操作流程【tg:seogoogle987】,iW7
- https://www.world-architects.com/en/search?term=谷歌快排操作流程【tg:seogoogle987】,kf7
- https://www.tmtpost.com/search?q=谷歌快排操作流程【tg:seogoogle987】,pDV
- https://www.reallywildbirdfood.co.uk/store/search?q=谷歌快排操作流程【tg:seogoogle987】,0yl
- https://www.otakuninjahero.com/en/search/谷歌快排操作流程【tg:seogoogle987】,H4S
- https://www.sensorshop24.de/catalogsearch/result/index/?type=132&q=谷歌快排操作流程【tg:seogoogle987】,GmL
- https://www.shopping.com/search.html?keyword=谷歌快排操作流程【tg:seogoogle987】,nJs
- https://www.seforall.org/search?keywords=谷歌快排操作流程【tg:seogoogle987】,nmU
- https://www.mygrandma.com/Items.aspx?search=谷歌快排操作流程【tg:seogoogle987】,lXa
- https://www.msrenewal.com/search/谷歌快排操作流程【tg:seogoogle987】,4rM
- https://www.remit.co.jp/en/?s=谷歌快排操作流程【tg:seogoogle987】,0E4
- https://www.rookwinkel.nl/catalogsearch/result/index/?q=谷歌快排操作流程【tg:seogoogle987】,Y2b
- https://www.samen-maier.at/search/谷歌快排操作流程【tg:seogoogle987】,jHa
- https://www.lssystems.co.uk/search?page=0&per_page=24&term=谷歌快排操作流程【tg:seogoogle987】,TsL
- https://www.kaiyueleather.com/search/谷歌快排操作流程【tg:seogoogle987】,UKq
- https://www.lezhin.com/ko/tags/谷歌快排操作流程【tg:seogoogle987】,IcC
- https://www.roberthalf.com/us/en/jobs/all/谷歌快排操作流程【tg:seogoogle987】XfY
- https://www.tcv.org.uk/search/谷歌快排操作流程【tg:seogoogle987】,ZMm
- https://www.abbreviations.com/谷歌快排操作流程【tg:seogoogle987】kMi
- https://www.mintos.org/?s=谷歌快排操作流程【tg:seogoogle987】,ReV
- https://www.kawainter.com/catalogsearch/result/?q=谷歌快排操作流程【tg:seogoogle987】,FjX
- https://www.sbaad.no/search/谷歌快排操作流程【tg:seogoogle987】,ZPw
- https://www.takachi-enclosure.com/search?eqYB=gnI31Ou&text=谷歌快排操作流程【tg:seogoogle987】,sPQ
- https://www.jusoen.com/addreng.asp?p1=谷歌快排操作流程【tg:seogoogle987】h6I
- https://www.shacknews.com/editorial-search?q=谷歌快排操作流程【tg:seogoogle987】,1Az
- https://www.voices.com/talents/search?keywords=谷歌快排操作流程【tg:seogoogle987】,BKh
- https://www.u-bourgogne.fr/?cat=117&s=谷歌快排操作流程【tg:seogoogle987】,GMs
- https://www.skicheck.at/en/?action=CHECK&s=谷歌快排操作流程【tg:seogoogle987】,w5i
- https://www.metalmensales.com/?action=QUERY&s=谷歌快排操作流程【tg:seogoogle987】,B3H
- https://www.vet-direct.com/search?cat=72&term=谷歌快排操作流程【tg:seogoogle987】,ZcA
- https://www.tumblr.com/tagged/谷歌快排操作流程【tg:seogoogle987】,LYh
- https://www.military-history.org/?s=谷歌快排操作流程【tg:seogoogle987】,sfG
- https://www.remit.co.jp/zh/?s=谷歌快排操作流程【tg:seogoogle987】,vd7
- https://www.repetto.jp/item/search?action=QUERY&q=谷歌快排操作流程【tg:seogoogle987】,iO3
- https://www.valuecolleges.com/?action=QUERY&s=谷歌快排操作流程【tg:seogoogle987】,JaR
- https://www.kaliastyle.com/catalogsearch/result/?q=GOOGLE谷歌快排操作流程【tg:seogoogle987】,xzq
- https://www.sic-info.org/en/?s=谷歌快排操作流程【tg:seogoogle987】,o8T
- https://www.simcoemuskokahealth.org/SearchResults.aspx?SearchQuery=谷歌快排操作流程【tg:seogoogle987】,2Ma
- https://xehay.vn/tim-kiem/=谷歌快排操作流程【tg:seogoogle987】,JKT
- https://5gaa.org/?token=6e2l7hjt2i26gzgctgwdrv7v0j6nxorq&s=谷歌快排操作流程【tg:seogoogle987】RlU
- https://www.mingwai.edu.hk/?langzh-hant&s=谷歌快排操作流程【tg:seogoogle987】,yXK
- https://www.synonyms.com/antonyms/谷歌快排操作流程【tg:seogoogle987】,ulP
- https://www.tullygroup.us/?ID=hbkq8jFjrQlVRrwzEmvoUJJAY4e1Lfo5&s=谷歌快排操作流程【tg:seogoogle987】,EU4
- https://www.luxurysociety.com/en/search?q=谷歌快排操作流程【tg:seogoogle987】,UAp
- https://www.qidulp.com/search?query=谷歌快排操作流程【tg:seogoogle987】,bSV
- https://www.ng-slo.si/si/iskanje?cat=114&q=谷歌快排操作流程【tg:seogoogle987】,mfk
- https://www.teacherspayteachers.com/Browse/Search:谷歌快排操作流程【tg:seogoogle987】,QN4
- https://www.toy-people.com/?s=谷歌快排操作流程【tg:seogoogle987】,vy2
- https://www.yr.no/en/search?q=谷歌快排操作流程【tg:seogoogle987】,wXl