修改成本地可以运行。
修改方法：

1.save_model.py|train.py|eval.py|run.py|中moxing.framework.file函数全部换成os.path和shutil.copy函数。因为python里面暂时没有moxing框架。

2.注释掉run.py文件里面的下面几行代码：

# FLAGS.tmp = os.path.join(FLAGS.local_data_root, 'tmp/')
# print(FLAGS.tmp)
# if not os.path.exists(FLAGS.tmp):

	#     os.mkdir(FLAGS.tmp)



	





	
	

		BaseLine改进
	

	

		1.使用多种模型进行对比实验，ResNet50, SE-ResNet50, Xeception, SE-Xeception, efficientNetB5。
	

	

		2.使用组归一化（GroupNormalization）代替批量归一化（batch_normalization）-解决当Batch_size过小导致的准确率下降。当batch_size小于16时，BN的error率
逐渐上升，train.py。
	

for i, layer in enumerate(model.layers):
    if "batch_normalization" in layer.name:
        model.layers[i] = GroupNormalization(groups=32, axis=-1, epsilon=0.00001)
	

		3.NAdam优化器
	
optimizer = Nadam(lr=FLAGS.learning_rate, beta_1=0.9, beta_2=0.999, epsilon=1e-08, schedule_decay=0.004)
	

		

			1
		
	
	

		4.自定义学习率-SGDR余弦退火学习率
	
sample_count = len(train_sequence) * FLAGS.batch_size
epochs = FLAGS.max_epochs
warmup_epoch = 5
batch_size = FLAGS.batch_size
learning_rate_base = FLAGS.learning_rate
total_steps = int(epochs * sample_count / batch_size)
warmup_steps = int(warmup_epoch * sample_count / batch_size)

warm_up_lr = WarmUpCosineDecayScheduler(learning_rate_base=learning_rate_base,
                                        total_steps=total_steps,
                                        warmup_learning_rate=0,
                                        warmup_steps=warmup_steps,
                                        hold_base_rate_steps=0,
                                        )
	

		5.数据增强：随机水平翻转、随机垂直翻转、以一定概率随机旋转90°、180°、270°、随机crop(0-10%)等(详细代码请看aug.py和data_gen.py)
	
def img_aug(self, img):
    data_gen = ImageDataGenerator()
    dic_parameter = {'flip_horizontal': random.choice([True, False]),
                     'flip_vertical': random.choice([True, False]),
                     'theta': random.choice([0, 0, 0, 90, 180, 270])
                    }


    img_aug = data_gen.apply_transform(img, transform_parameters=dic_parameter)
    return img_aug


from imgaug import augmenters as iaa
import imgaug as ia

def augumentor(image):
    sometimes = lambda aug: iaa.Sometimes(0.5, aug)
    seq = iaa.Sequential(
        [
            iaa.Fliplr(0.5),
            iaa.Flipud(0.5),
            iaa.Affine(rotate=(-10, 10)),
            sometimes(iaa.Crop(percent=(0, 0.1), keep_size=True)),
        ],
        random_order=True
    )


    image_aug = seq.augment_image(image)

    return image_aug
	

		6.标签平滑data_gen.py
	
def smooth_labels(y, smooth_factor=0.1):
    assert len(y.shape) == 2
    if 0 <= smooth_factor <= 1:
        # label smoothing ref: https://www.robots.ox.ac.uk/~vgg/rg/papers/reinception.pdf
        y *= 1 - smooth_factor
        y += smooth_factor / y.shape[1]
    else:
        raise Exception(
            'Invalid label smoothing factor: ' + str(smooth_factor))
    return y