Created
December 13, 2021 18:49
-
-
Save wbuchwalter/a08a6a24bef9c5370ba312cf340ea14a to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import numpy as np | |
from os import listdir | |
from os.path import isfile, isdir, join | |
import os | |
import json | |
import random | |
import re | |
cwd = os.getcwd() | |
cwd = os.getcwd() | |
datadir = cwd.split('filelists')[0] | |
data_path = join(datadir,'Datasets/ILSVRC/Data/CLS-LOC/') | |
savedir = './' | |
dataset_list = ['base', 'val', 'novel'] | |
cl = -1 | |
folderlist = [] | |
datasetmap = {'base':'train','val':'val','novel':'test'}; | |
filelists = {'base':{},'val':{},'novel':{} } | |
filelists_flat = {'base':[],'val':[],'novel':[] } | |
labellists_flat = {'base':[],'val':[],'novel':[] } | |
for dataset in dataset_list: | |
with open(datasetmap[dataset] + ".csv", "r") as lines: | |
for i, line in enumerate(lines): | |
if i == 0: | |
continue | |
fid, _ , label = re.split(',|\.', line) | |
label = label.replace('\n','') | |
if not label in filelists[dataset]: | |
folderlist.append(label) | |
filelists[dataset][label] = [] | |
fnames = listdir( join(data_path, datasetmap[dataset], label) ) | |
fname_number = [ int(re.split('_|\.', fname)[0][1:]) for fname in fnames] | |
sorted_fnames = list(zip( *sorted( zip(fnames, fname_number), key = lambda f_tuple: f_tuple[1] )))[0] | |
print('sorting') | |
# import pdb; pdb.set_trace() | |
#fid2 = int(fid[-5:])-1 | |
s = sorted_fnames[i%600-1] | |
if fid+'.jpg' != s: | |
print('uneq') | |
import pdb; pdb.set_trace() | |
fname = join( data_path,label, sorted_fnames[i%600-1] ) | |
filelists[dataset][label].append(fname) | |
for key, filelist in filelists[dataset].items(): | |
cl += 1 | |
random.shuffle(filelist) | |
filelists_flat[dataset] += filelist | |
labellists_flat[dataset] += np.repeat(cl, len(filelist)).tolist() | |
for dataset in dataset_list: | |
fo = open(savedir + dataset + ".json", "w") | |
fo.write('{"label_names": [') | |
fo.writelines(['"%s",' % item for item in folderlist]) | |
fo.seek(0, os.SEEK_END) | |
fo.seek(fo.tell()-1, os.SEEK_SET) | |
fo.write('],') | |
fo.write('"image_names": [') | |
fo.writelines(['"%s",' % item for item in filelists_flat[dataset]]) | |
fo.seek(0, os.SEEK_END) | |
fo.seek(fo.tell()-1, os.SEEK_SET) | |
fo.write('],') | |
fo.write('"image_labels": [') | |
fo.writelines(['%d,' % item for item in labellists_flat[dataset]]) | |
fo.seek(0, os.SEEK_END) | |
fo.seek(fo.tell()-1, os.SEEK_SET) | |
fo.write(']}') | |
fo.close() | |
print("%s -OK" %dataset) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment