import gradio as gr
from transformers import pipeline
import numpy as np
from PIL import Image

pipe = pipeline("zero-shot-image-classification", model="openai/clip-vit-base-patch32")

def shot(image, labels_text):
    PIL_image = Image.fromarray(np.uint8(image)).convert('RGB')
    labels = labels_text.split(";;")
    res = pipe(images=PIL_image, 
           candidate_labels=labels,
           hypothesis_template="This is a photo of {}")
    return {dic["label"]: dic["score"] for dic in res}
    
iface = gr.Interface(shot, 
                    ["image", "text"], 
                    "label", 
                    examples=[
							  ["examples/1.jpg", "ralph lauren;;apparel store;;ralph lauren store;;shirts;;wardrobe;;white flower"], 
                              ["examples/2.JPG", "adidas;;apparel store;;adidas store;;shirts;;wardrobe;;women training;;shoes"], 
                              ["examples/3.jpg", "project x;;sweet monster;;bags store;;store;;shoes store;;glass windows;;hanging lights"], 
                              ["examples/4.JPG", "multi brand store;;multi brand shoe store;;shoe store;;mannequins;;adidas store;;reebok store;;puma store"], 
                              ["examples/5.png", "sophie;;scene"],
							 ],
                    description="Add a picture and a list of labels separated by ;;",
                    title="Zero-shot Image Classification")

iface.launch()