I have my dataset images in a folder in .jpg format and the labels in a csv file
I load the images into a tensor
by
dname = paths.dirname('/home/soundar/Videos/Train/Train_Data/5ChildrenAndIt_audio_00014_non-human1.jpg')
count = 0
for a in paths.files(dname) do
if a:find('jpg') then
count = count+1
end
end
print(count)
train_set = torch.Tensor(count, 3, 96, 96):float()
count = 0
for a in paths.files(dname) do
if a:find('jpg') then
collectgarbage()
count = count + 1
a = dname .. "/" .. a
train_set[count] = image.load(a)
print(count)
end
end
when I train I always get true negative and false positive zero,
I doubt whether the images are loaded in the same order as in the folder because the labels are loaded in order.
Can anyone tell me whether they will be loaded in order or some other way to load in order