6 自制Fashion数据集

最新推荐文章于 2024-11-14 17:31:27 发布

CourageLee

最新推荐文章于 2024-11-14 17:31:27 发布

阅读量401

点赞数

分类专栏： tensorflow 文章标签： tensorflow

本文链接：https://blog.csdn.net/qq_35779738/article/details/110224936

版权

本文将指导你如何利用个人收集的资料创建自己的Fashion数据集，适用于深度学习模型训练，特别是TensorFlow项目。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

学习如何根据自己已有的数据制作数据集！

import  os
from  PIL import  Image
import numpy as np
import tensorflow as tf
from tensorflow.keras.models import Model
from  tensorflow.keras.layers import  Flatten ,Dense,Conv1D,MaxPool1D
from tensorflow.keras.preprocessing.image import  ImageDataGenerator
import matplotlib.pyplot as plt

x_train_path = './data/class4/FASHION_FC/fashion_image_label/fashion_train_jpg_60000/'
y_train_path = './data/class4/FASHION_FC/fashion_image_label/fashion_train_jpg_60000.txt'
x_test_path = './data/class4/FASHION_FC/fashion_image_label/fashion_test_jpg_10000/'
y_test_path = './data/class4/FASHION_FC/fashion_image_label/fashion_test_jpg_10000.txt'

x_train_savepath = './data/class4/FASHION_FC/fashion_image_label/fashion_x_train.npy'
y_train_savepath = './data/class4/FASHION_FC/fashion_image_label/fashion_y_train.npy'
x_test_savepath = './data/class4/FASHION_FC/fashion_image_label/fashion_x_test.npy'
y_test_savepath = './data/class4/FASHION_FC/fashion_image_label/fashion_y_test.npy'

def generateds(path,txt):
    file = open(txt,'r')
    lines = file.readlines()
    file.close()
    x,y_hat = [],[]
    for line in lines:
        value = line.split()
        img_path = path +value[0]
        img = Image.open(img_path)
        img = np.array(img.convert('L'))
        img = img /255.0

        x.append(img)
        y_hat.append(value[1])
        print('loading'+line

最低0.47元/天解锁文章