app.py

import torch
from torchvision import models, transforms
from PIL import Image


labels = {
  0: "bluebell",
  1: "buttercup",
  2: "colts_foot",
  3: "corn_poppy",
  4: "cowslip",
  5: "crocus",
  6: "daffodil",
  7: "daisy",
  8: "dandelion",
  9: "foxglove",
  10: "fritillary",
  11: "geranium",
  12: "hibiscus",
  13: "iris",
  14: "lily_valley",
  15: "pansy",
  16: "petunia",
  17: "rose",
  18: "snowdrop",
  19: "sunflower",
  20: "tigerlily",
  21: "tulip",
  22: "wallflower",
  23: "water_lily",
  24: "wild_tulip",
  25: "windflower"
}

# Load the trained ResNet-152 model
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

# Load model structure
model = models.resnet152()
num_classes = 26  # Update with your dataset's number of classes
model.fc = torch.nn.Linear(model.fc.in_features, num_classes)

# Load trained weights
model.load_state_dict(torch.load('trained_model.pth', map_location=device))
model = model.to(device)
model.eval()  # Set to evaluation mode

# Preprocessing pipeline for incoming images
preprocess = transforms.Compose([
    transforms.Resize((224, 224)),  # ResNet default input size
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
])

def predict_image(image_path):
    # Load and preprocess the image
    image = Image.open(image_path).convert("RGB")
    input_tensor = preprocess(image).unsqueeze(0).to(device)

    # Predict
    with torch.no_grad():
        outputs = model(input_tensor)
        _, predicted_class = torch.max(outputs, 1)

    return predicted_class.item()  # Return class index

import gradio as gr

def get_class_name(class_index):
    return labels[class_index]
# Function to predict from an uploaded image
def classify_image(image):
    predicted_class = predict_image(image)  # Use the function from above
    return f"Predicted Class: {predicted_class} : {get_class_name(predicted_class)}"

# Create Gradio interface
interface = gr.Interface(
    fn=classify_image,
    inputs=gr.Image(type="filepath"),  # Accept image uploads
    outputs="text",
    title="Image Classification with ResNet-152",
    description="Upload an image to classify it into one of 26 classes."
)

# Launch the app
interface.launch()