app.py

import torch
from transformers import DetrForObjectDetection, DetrImageProcessor
from PIL import Image
import gradio as gr
import matplotlib.pyplot as plt
import matplotlib.patches as patches

# Load pre-trained model and processor
model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50")
processor = DetrImageProcessor.from_pretrained("facebook/detr-resnet-50")

def detect_car(image: Image.Image) -> Image.Image:
    # Preprocess the input image
    inputs = processor(images=image, return_tensors="pt")

    # Run the model to get predictions
    outputs = model(**inputs)

    # Postprocess the outputs to get bounding boxes and labels
    target_sizes = torch.tensor([image.size[::-1]])  # (height, width)
    results = processor.post_process_object_detection(outputs, target_sizes=target_sizes)[0]

    # Plotting the image with bounding boxes for objects
    fig, ax = plt.subplots(1, figsize=(12, 8))
    ax.imshow(image)

    for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
        if score > 0.7:  # Confidence threshold for detecting cars
            xmin, ymin, xmax, ymax = box.detach().numpy()
            width, height = xmax - xmin, ymax - ymin
            rect = patches.Rectangle((xmin, ymin), width, height, linewidth=2, edgecolor='red', facecolor='none')
            ax.add_patch(rect)
            ax.text(xmin, ymin, f"{model.config.id2label[label.item()]}: {score:.2f}",
                    color='white', fontsize=12, bbox=dict(facecolor='red', alpha=0.5))

    # Convert the plot to an image
    plt.axis('off')
    plt.tight_layout()

    # Save the figure to a canvas and convert to image
    fig.canvas.draw()
    result_img = Image.frombytes('RGB', fig.canvas.get_width_height(), fig.canvas.tostring_rgb())
    plt.close(fig)
    return result_img

# Gradio interface to upload images and get object detection results
iface = gr.Interface(
    fn=detect_car,
    inputs=gr.Image(type="pil"),
    outputs=gr.Image(type="pil"),
    title="Car Detection with DETR",
    description="Upload an image and the model will detect cars with bounding boxes. Only cars will be displayed."
)

if __name__ == "__main__":
    iface.launch()