app.py

import gradio as gr

from torchvision import models
import torch.nn as nn
import torch
import os
from PIL import Image
from torchvision.transforms import transforms
from dotenv import load_dotenv
load_dotenv()

share = os.getenv("SHARE", False)
pretrained_model = models.vgg19(pretrained=True)
class NeuralNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.model = nn.Sequential(
            pretrained_model,
            nn.Flatten(),
            nn.Linear(1000, 1),
            nn.Sigmoid()
        )

    def forward(self, x):
        return self.model(x)

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

model = NeuralNet()

model.load_state_dict(torch.load("mask_detection.pth", map_location=device))

model = model.to(device)

transform=transforms.Compose([
    transforms.Resize((150,150)),
    transforms.RandomHorizontalFlip(),
    transforms.ToTensor(),
    transforms.Normalize([0.5, 0.5, 0.5],[0.5, 0.5, 0.5])
])

def greet(image):
    image = Image.fromarray(image.astype('uint8'), 'RGB')
    image.save("input.png")
    image = Image.open("input.png")
    input = transform(image).unsqueeze(0)
    output = model(input.to(device))
    probability = output.item()
    if probability < 0.5:
        return "Person in the pic has mask"
    else:
        return "Person in the pic does not have mask"

iface = gr.Interface(fn=greet, inputs="image", outputs="text")
iface.launch()