aliwaqas333

"value": {
              "format": "rle",
              "rle": [
                0, 18, 192, 0, 57, 27, 255, 255, 255, 0, 255, 255, 224, 31, 255,
                252, 3, 255, 255, 128, 113, 224, 112, 8, 212, 35, 207, 159, 254,
                58, 248, 211, 248, 79, 24, 4, 100, 24, 95, 254, 61, 248, 223,
                227, 16, 225, 58, 224, 17, 168, 97, 255, 248, 251, 227, 80, 225,
                58, 96, 17, 168, 98, 127, 248, 239, 227, 64, 225, 57, 96, 17,
                144, 98, 255, 248, 239, 227, 100, 140, 255, 132, 225, 128, 70,
                161, 30, 134, 51, 255, 143, 190, 52, 14, 19, 142, 1, 27, 4, 120,

aliwaqas333

#!/bin/bash

cd ~/Desktop

# ---------------------------
## 0. Requirements
# ---------------------------
sudo apt-get update
sudo apt-get install -y build-essential git pkg-config cmake make \
    gcc curl wget unzip \

aliwaqas333

def singleImage(path, label= None, show= False):
    img = cv2.imread(path)
    assert img is not None,"Immage wasn't read properly"
    img = cv2.resize(img, (100, 100))
    img = torch.from_numpy(img)
    img = img.permute((2, 0,1)) # model expects image to be of shape [3, 100, 100]
    img = img.unsqueeze(dim=0).float() # convert single image to batch [1, 3, 100, 100]
    img = img.to('cuda') # Using the same device as the model
    pred = model(img)
    _, preds = torch.max(pred, dim=1)

aliwaqas333

@torch.no_grad()
def evaluate(model, val_loader):
    model.eval()
    outputs = [model.validation_step(batch) for batch in val_loader]
    return model.validation_epoch_end(outputs)

def fit(epochs, lr, model, train_loader, val_loader, opt_func=torch.optim.SGD):
    history = []
    optimizer = torch.optim.SGD(model.parameters(), lr=0.001, momentum=0.9)
    for epoch in range(epochs):

aliwaqas333

def get_default_device():
    """Pick GPU if available, else CPU"""
    if torch.cuda.is_available():
        return torch.device('cuda')
    else:
        return torch.device('cpu')

device = get_default_device()

def to_device(data, device):

aliwaqas333

class ImageClassificationBase(nn.Module):
    def training_step(self, batch):
        images, labels = batch 
        out = self(images)                  # Generate predictions
        loss = F.cross_entropy(out, labels.long()) # Calculate loss
        return loss
    
    def validation_step(self, batch):
        images, labels = batch 
        out = self(images)                    # Generate predictions

aliwaqas333

class MaskDetection(ImageClassificationBase):
    def __init__(self):
        super().__init__()
        self.network = nn.Sequential(
            nn.Conv2d(3, 100, kernel_size=3, padding=1),
            nn.ReLU(),
            nn.Conv2d(100, 128, kernel_size=3, stride=1, padding=1),
            nn.ReLU(),
            nn.MaxPool2d(2, 2), # output: 128 x 8 x 8

aliwaqas333

from torch.utils.data.dataloader import DataLoader
from torchvision.utils import make_grid
batch_size = 32

train_dl = DataLoader(train_ds, batch_size*2, shuffle=True)
val_dl = DataLoader(val_ds, batch_size*2)

aliwaqas333

from torch.utils.data.dataset import Dataset
from torchvision.transforms import Compose, ToTensor
class MaskDataset(Dataset):
        """ Masked faces dataset
        0 = 'no mask'
        1 = 'mask'
        """
        def __init__(self, train_data):
            self.train_data = train_data
            self.transformations = Compose([

aliwaqas333

REBUILD_DATA = True

if REBUILD_DATA: #if we are running it for the first time
    data_path = Path('F:/FILES/AI/face-mask-dataset/')
    maskPath = data_path/'dataset1/AFDB_masked_face_dataset'
    maskPath2= data_path/'dataset2/webface_masked'
    nonMaskPath = data_path/'dataset1/AFDB_face_dataset'
    path_dirs = [ [maskPath,1],[nonMaskPath,0] ] #path and label
    if not os.path.exists(data_path):
        raise Exception("The data path doesn't exist")