How to make imaged based authentication?

Using Streamlit
, ,
1

Hello I want to make a a simple page where the user can upload an image, then that image is fed to a model, if the confidence of the model is higher than a certain threshold (for ex: 97%) then the user gets authenticated. How do I implement this using streamlit?

2

To perform inference locally, you will need to utilize a machine learning library. Alternatively, you can employ an API for a hosted model. Below is an example utilizing PyTorch:

import streamlit as st
from PIL import Image
import torch
from torchvision import transforms

# Load a pre-trained model (e.g., ResNet) for image classification
model = torch.hub.load('pytorch/vision', 'resnet18', pretrained=True)
model.eval()

# Define image transformation
transform = transforms.Compose([
    transforms.Resize(256),
    transforms.CenterCrop(224),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
])

# Function to predict image class
def predict_image(image):
    image = transform(image).unsqueeze(0)  # Transform image and add batch dimension
    with torch.no_grad():
        outputs = model(image)  # Forward pass
        probs = torch.nn.functional.softmax(outputs, dim=1)
        confidence, predicted_class = probs.max(1)
    return confidence.item(), predicted_class.item()

# Streamlit UI
st.title("Image Authentication")

# Upload image
uploaded_file = st.file_uploader("Choose an image", type=["jpg", "png", "jpeg"])

if uploaded_file is not None:
    # Display image
    image = Image.open(uploaded_file)
    st.image(image, caption="Uploaded Image", use_column_width=True)

    # Get prediction
    confidence, predicted_class = predict_image(image)

    # Set a confidence threshold (e.g., 97%)
    threshold = 0.97

    # Display result
    if confidence >= threshold:
        st.success(f"Authentication successful! Confidence: {confidence * 100:.2f}%")
    else:
        st.error(f"Authentication failed. Confidence: {confidence * 100:.2f}%")
1 Like
3

this isn’t secure, I need to authenticate the user to a secret page

4

To implement multiple protected pages with authentication, you can utilize the code example provided in the following GitHub repository.