app.py

# !pip install ultralytics gradio supervision torch pillow opencv-python
import spaces
import supervision as sv
import PIL.Image as Image
import cv2
import numpy as np
from ultralytics import YOLO
import gradio as gr
import torch

# YOLO model filenames
model_filenames = [
    "yolo11n.pt",
    "yolo11s.pt",
    "yolo11m.pt",
    "yolo11l.pt",
    "yolo11x.pt"
]

# Box annotator for drawing bounding boxes
box_annotator = sv.BoxAnnotator()

# COCO category dictionary for labeling classes
category_dict = {
    0: 'person', 1: 'bicycle', 2: 'car', 3: 'motorcycle', 4: 'airplane', 5: 'bus',
    6: 'train', 7: 'truck', 8: 'boat', 9: 'traffic light', 10: 'fire hydrant',
    11: 'stop sign', 12: 'parking meter', 13: 'bench', 14: 'bird', 15: 'cat',
    16: 'dog', 17: 'horse', 18: 'sheep', 19: 'cow', 20: 'elephant', 21: 'bear',
    22: 'zebra', 23: 'giraffe', 24: 'backpack', 25: 'umbrella', 26: 'handbag',
    27: 'tie', 28: 'suitcase', 29: 'frisbee', 30: 'skis', 31: 'snowboard',
    32: 'sports ball', 33: 'kite', 34: 'baseball bat', 35: 'baseball glove',
    36: 'skateboard', 37: 'surfboard', 38: 'tennis racket', 39: 'bottle',
    40: 'wine glass', 41: 'cup', 42: 'fork', 43: 'knife', 44: 'spoon', 45: 'bowl',
    46: 'banana', 47: 'apple', 48: 'sandwich', 49: 'orange', 50: 'broccoli',
    51: 'carrot', 52: 'hot dog', 53: 'pizza', 54: 'donut', 55: 'cake',
    56: 'chair', 57: 'couch', 58: 'potted plant', 59: 'bed', 60: 'dining table',
    61: 'toilet', 62: 'tv', 63: 'laptop', 64: 'mouse', 65: 'remote', 66: 'keyboard',
    67: 'cell phone', 68: 'microwave', 69: 'oven', 70: 'toaster', 71: 'sink',
    72: 'refrigerator', 73: 'book', 74: 'clock', 75: 'vase', 76: 'scissors',
    77: 'teddy bear', 78: 'hair drier', 79: 'toothbrush'
}

# YOLO inference function
@spaces.GPU
def yolo_inference(image, model_id, conf_threshold, iou_threshold, max_detection):
    model = YOLO(model_id)
    # model.to("cuda")
    results = model(source=image, imgsz=640, iou=iou_threshold, conf=conf_threshold, verbose=False, max_det=max_detection)[0]
    detections = sv.Detections.from_ultralytics(results)

    # Count objects
    counts = {}
    for class_id in detections.class_id:
        label = category_dict[class_id]
        if label not in counts:
            counts[label] = 0
        counts[label] += 1

    # Prepare labels for drawing boxes and counting
    labels = [
        f"{category_dict[class_id]} {confidence:.2f}"
        for class_id, confidence in zip(detections.class_id, detections.confidence)
    ]

    # Annotate the image with bounding boxes
    annotated_image = box_annotator.annotate(image, detections=detections, labels=labels)

    # Convert annotated_image to OpenCV format (from PIL)
    annotated_image_cv = cv2.cvtColor(np.array(annotated_image), cv2.COLOR_RGB2BGR)

    # Draw counts on the annotated image using cv2.putText
    y_offset = 30  # Starting y offset for text
    for label, count in counts.items():
        text = f"{label}: {count}"
        cv2.putText(annotated_image_cv, text, (10, y_offset), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0, 0, 255), 2, cv2.LINE_AA)
        y_offset += 25  # Increase y offset for the next label

    # Convert back to PIL for Gradio output
    return Image.fromarray(cv2.cvtColor(annotated_image_cv, cv2.COLOR_BGR2RGB))

# Gradio app function
def app():
    with gr.Blocks():
        with gr.Row():
            with gr.Column():
                image = gr.Image(type="pil", label="Image", interactive=True)
                model_id = gr.Dropdown(label="Model", choices=model_filenames, value=model_filenames[0] if model_filenames else "")
                conf_threshold = gr.Slider(label="Confidence Threshold", minimum=0.1, maximum=1.0, step=0.1, value=0.25)
                iou_threshold = gr.Slider(label="IoU Threshold", minimum=0.1, maximum=1.0, step=0.1, value=0.45)
                max_detection = gr.Slider(label="Max Detection", minimum=1, maximum=300, step=1, value=300)
                yolov_infer = gr.Button(value="Detect Objects")
            with gr.Column():
                output_image = gr.Image(type="pil", label="Annotated Image", interactive=False)
        yolov_infer.click(fn=yolo_inference, inputs=[image, model_id, conf_threshold, iou_threshold, max_detection], outputs=[output_image])

# Main Gradio app
gradio_app = gr.Blocks()
with gradio_app:
    gr.HTML("<h1 style='text-align: center'>Object Counting using YoloV11</h1>")
    gr.HTML("<p style='text-align: center'>Upload an image to run inference. By Kelvin</p>")
    with gr.Row():
        with gr.Column():
            app()

# gradio_app.launch(debug=True)
gradio_app.launch()