[Image] add random lighting

[sxjscience]

Description

Add RandomLighting and AdjustLighting
#8556

import os
os.environ['MXNET_CPU_WORKER_NTHREADS'] = '1'
os.environ['OMP_NUM_THREADS'] = '1'
import time
import numpy as np
import multiprocessing as mp
import mxnet as mx
from mxnet import gluon as gl
from mxnet.gluon.data.vision import transforms

if __name__ == '__main__':
    M = 24
    BS = 100

    dataset = gl.data.vision.ImageFolderDataset('256_ObjectCategories')
    transform = transforms.Compose([transforms.RandomBrightness(1.0),
                                    transforms.RandomContrast(1.0),
                                    transforms.RandomSaturation(1.0),
                                    transforms.RandomLighting(0.05),
                                    transforms.ToTensor(),
                                    transforms.Normalize([0, 0, 0], [1, 1, 1])])
    dataset = dataset.transform_first(lambda x: transform(mx.image.center_crop(x, (224, 224))[0]))
    data_loader = gl.data.DataLoader(dataset, BS, shuffle=True, num_workers=M)

    N = len(dataset)

    iterator = iter(data_loader)

    tic = time.time()

    for data, label in iterator:
        data.wait_to_read()
        print(data.shape)

    print(N/(time.time() - tic))

[piiswrong]

don't need this.

[sxjscience]

OK. Then I will directly call the operator in the testing code

[piiswrong]

what's alpha_rgb?
use base._Null for default arguments to avoid parsing

[sxjscience]

It's (alpha_r, alpha_g, alpha_b)

[piiswrong]

merge to one loop