def __init__(self, root, batch_size, train=False, input_size=224, **kwargs):
self.mean = np.array([0.485, 0.456, 0.406]).reshape(1, 1, 1, 3)
self.std = np.array([0.229, 0.224, 0.225]).reshape(1, 1, 1, 3)
self.train = train
if train:
pkl_file = os.path.join(root, 'train{}.pkl'.format(input_size))
else:
pkl_file = os.path.join(root, 'val{}.pkl'.format(input_size))
self.data_dict = misc.load_pickle(pkl_file)
self.batch_size = batch_size
self.idx = 0
from utee import misc
import argparse
import cv2
imagenet_urls = [
'http://ml.cs.tsinghua.edu.cn/~chenxi/dataset/val224_compressed.pkl'
]
parser = argparse.ArgumentParser(
description='Extract the ILSVRC2012 val dataset')
parser.add_argument('--in_file',
default='./val224_compressed.pkl',
help='input file path')
parser.add_argument('--out_root', default='~/dataset', help='output file path')
args = parser.parse_args()
d = misc.load_pickle(args.in_file)
assert len(d['data']) == 50000, len(d['data'])
assert len(d['target']) == 50000, len(d['target'])
data224 = []
data299 = []
for img, target in tqdm.tqdm(zip(d['data'], d['target']), total=1000):
img224 = misc.str2img(img)
#img299 = cv2.resize(img224, (299, 299))
data224.append(img224)
#data299.append(img299)
data_dict224 = dict(data=np.array(data224).transpose(0, 3, 1, 2),
target=d['target'])
#data_dict299 = dict(
# data = np.array(data299).transpose(0, 3, 1, 2),
import os
import numpy as np
import tqdm
from utee import misc
import argparse
import cv2
imagenet_urls = [
'http://ml.cs.tsinghua.edu.cn/~chenxi/dataset/val224_compressed.pkl'
]
parser = argparse.ArgumentParser(description='Extract the ILSVRC2012 val dataset')
parser.add_argument('--in_file', default='val224_compressed.pkl', help='input file path')
parser.add_argument('--out_root', default='/tmp/public_dataset/pytorch/imagenet-data/', help='output file path')
args = parser.parse_args()
d = misc.load_pickle(args.in_file)
assert len(d['data']) == 50000, len(d['data'])
assert len(d['target']) == 50000, len(d['target'])
data224 = []
data299 = []
for img, target in tqdm.tqdm(zip(d['data'], d['target']), total=50000):
img224 = misc.str2img(img)
img299 = cv2.resize(img224, (299, 299))
data224.append(img224)
data299.append(img299)
data_dict224 = dict(
data = np.array(data224).transpose(0, 3, 1, 2),
target = d['target']
)
data_dict299 = dict(
#image.save("test_1.png", 'png')
print(data0_to3)
print("+++++")
print(np.array(data0_to3).transpose(1, 2, 0))
# xx = -1/3 * math.log2(1/3) - 2/3 * math.log2(2/3)
# xx1 = -1/5 * math.log2(1/5) - 4/5 * math.log2(4/5)
# print(xx,xx1)
from utee import misc
from collections import Counter
pkl_path = "../tmp/public_dataset/pytorch/imagenet-data/"
d = misc.load_pickle(pkl_path + 'val224.pkl')
data = d['data']
target = d['target']
print(len(data), len(target))
print(data[1].shape)
print(target)
result = Counter(target)
print(result)
for index, item in enumerate(data):
if target[index] == 823:
print("---")
dt = np.array(item).transpose(1, 2, 0)
cv2.imwrite("imagenet_" + str(index) + ".png", dt)
# data2 = np.array(data[5]).transpose(1, 2, 0)