app.py

from collections import defaultdict

import gradio as gr
import PIL.Image as Image
from ultralytics import YOLO

model = YOLO("best.pt")


def predict_image(img, conf_threshold, iou_threshold):
    """Predicts and plots labeled objects in an image using YOLOv8 model with adjustable confidence and IOU thresholds."""
    results = model.predict(
        source=img,
        conf=conf_threshold,
        iou=iou_threshold,
        show_labels=True,
        show_conf=True,
        imgsz=640,
    )

    label_counts = defaultdict(int)
    for r in results:
        if r.boxes is not None:
            for box in r.boxes:
                label = r.names[
                    int(box.cls.item())
                ]  # Convert tensor to int and then get the label
                label_counts[label] += 1
        im_array = r.plot()
        im = Image.fromarray(im_array[..., ::-1])

    label_counts_str = "\n".join(
        [f"{label}: {count}" for label, count in label_counts.items()]
    )
    return im, label_counts_str


logo_html = """
<style>
    table {
        margin-left: auto;
        margin-right: auto;
    }
</style>
<table class="table">
<tbody>
<tr>
<td><img width="100px" src="https://upload.wikimedia.org/wikipedia/commons/e/eb/Univalle.svg" alt="Univalle"> </td>
<td><img width="220px" src="https://i.ibb.co/6vdWxb4/PSI-LOGO.png" alt="PSI"></td>
</tr>
</tbody>
</table>
<h1>Reconocimiento de Medicamentos</h1>
<p>Subir Imagenes para Inferencia con un modelo customizado de YOLOv8</p>
<b>Creado Por:</b></br>
Edwar Stiven Montaño Cely </br>
<b>Correo:</b> Edwar.montano@correounivalle.edu.co</br>
<h2>Imágenes de las 10 cajas de medicamentos seleccionados para el estudio</h2>
<img
  sizes="(max-width: 600px) 480px,
         800px"
  src="https://pub-ae3fb8dcc56d4bcb9396023dc3901a9b.r2.dev/Medicamentos%20a%20reconocer.png"
  alt="Medicamentos" />
"""

iface = gr.Interface(
    fn=predict_image,
    inputs=[
        gr.Image(type="pil", label="Upload Image"),
        gr.Slider(minimum=0, maximum=1, value=0.4, label="Confidence threshold"),
        gr.Slider(minimum=0, maximum=1, value=0.5, label="IoU threshold"),
    ],
    outputs=[
        gr.Image(type="pil", label="Result"),
        gr.Textbox(label="Total Labels Recognized"),
    ],
    title="Reconocimiento de  Mediacamentos",
    description=logo_html,
    examples=[
        ["assets/01.jpeg", 0.38, 0.45],
        ["assets/02.jpeg", 0.38, 0.45],
        ["assets/03.jpeg", 0.4, 0.5],
        ["assets/04.jpeg", 0.4, 0.5],
    ],
    allow_flagging="never",
)

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