mk-gurucharan / face-mask-detection Goto Github PK

Great work!!! Is there any way to deploy it on mobile

what would be the perameter of this fun
def split_data(SOURCE, TRAINING, TESTING, SPLIT_SIZE):
dataset = []

for unitData in os.listdir(SOURCE):
    data = SOURCE + unitData
    if(os.path.getsize(data) > 0):
        dataset.append(unitData)
    else:
        print('Skipped ' + unitData)
        print('Invalid file i.e zero size')

train_set_length = int(len(dataset) * SPLIT_SIZE)
test_set_length = int(len(dataset) - train_set_length)
shuffled_set = random.sample(dataset, len(dataset))
train_set = dataset[0:train_set_length]
test_set = dataset[-test_set_length:]
   
for unitData in train_set:
    temp_train_set = SOURCE + unitData
    final_train_set = TRAINING + unitData
    copyfile(temp_train_set, final_train_set)

for unitData in test_set:
    temp_test_set = SOURCE + unitData
    final_test_set = TESTING + unitData
    copyfile(temp_test_set, final_test_set)

YES_SOURCE_DIR = "dest_folder/val/yes/"
TRAINING_YES_DIR = "dest_folder/train/yes/"
TESTING_YES_DIR = "dest_folder/test/yes/"
NO_SOURCE_DIR = "dest_folder/val/no/"
TRAINING_NO_DIR = "dest_folder/val/no/"
TESTING_NO_DIR = "dest_folder/test/no/"
split_size = .8
split_data(YES_SOURCE_DIR, TRAINING_YES_DIR, TESTING_YES_DIR, split_size)
split_data(NO_SOURCE_DIR, TRAINING_NO_DIR, TESTING_NO_DIR, split_size)

Hi Gurucharan
I have a question when I detect facemask.
It only show red rectangle “without_mask” when i don't wear a mask.
It cann't identify when I already wear a mask.
I hope you can give me assistance.
Thank you.

getting error in this line of code "mini = cv2.resize(im, (im.shape[1] // size, im.shape[0] // size))".
Would you please elaborate about "im"?What it is?

How you had uploaded dataset? Or how can I upload dataset on google colab?