app.py

import gradio as gr
import torch
from PIL import Image
from  story import story_model

# Images
torch.hub.download_url_to_file('https://github.com/ultralytics/yolov5/raw/master/data/images/zidane.jpg', 'zidane.jpg')
torch.hub.download_url_to_file('https://github.com/ultralytics/yolov5/raw/master/data/images/bus.jpg', 'bus.jpg')

# Model
model = torch.hub.load('ultralytics/yolov5', 'yolov5s')  # force_reload=True to update


def yolo(im, size=640):
    g = (size / max(im.size))  # gain
    im = im.resize((int(x * g) for x in im.size), Image.ANTIALIAS)  # resize

    results = model(im)  # inference

    r= results.pred[0].numpy().T
    #probs = results.pred[0].numpy().T[-1]
    res = [(results.names[x].lower(),p) for x,p in zip(r[-1].astype(int),r[-2])]
    f_res  = story_model(res)
    #results.render()  # updates results.imgs with boxes and labels
    #return Image.fromarray(results.imgs[0])
    return ','.join(f_res)


inputs = gr.inputs.Image(type='pil', label="Original Image")
#outputs = gr.outputs.Image(type="pil", label="Output Image")
outputs = gr.outputs.Textbox(type="str", label="Output Story")

title = "YOLOv5"
description = "YOLOv5 Gradio demo for object detection. Upload an image or click an example image to use."
article = "<p style='text-align: center'>YOLOv5 is a family of compound-scaled object detection models trained on the COCO dataset, and includes simple functionality for Test Time Augmentation (TTA), model ensembling, hyperparameter evolution, and export to ONNX, CoreML and TFLite. <a href='https://github.com/ultralytics/yolov5'>Source code</a> | <a href='https://pytorch.org/hub/ultralytics_yolov5'>PyTorch Hub</a></p>"

examples = [['zidane.jpg'], ['bus.jpg']]
iface = gr.Interface(yolo, inputs, outputs, title=title, description=description, article=article, examples=examples, theme="huggingface").launch(cache_examples=True,enable_queue=True)
iface.launch()