app.py

from PIL import Image, ImageFilter
import numpy as np
from transformers import pipeline
import gradio as gr
import os

models = [
    "facebook/detr-resnet-50-panoptic",
    "CIDAS/clipseg-rd64-refined",
    "facebook/maskformer-swin-large-ade",
    "nvidia/segformer-b1-finetuned-cityscapes-1024-1024",    
]
current_model = models[0]



#model = pipeline("image-segmentation", model="facebook/detr-resnet-50-panoptic")

pred = []

def img_resize(image):
    width = 1280
    width_percent = (width / float(image.size[0]))
    height = int((float(image.size[1]) * float(width_percent)))
    return image.resize((width, height))

def image_objects(image):
    global pred
    
    image = img_resize(image)
    pred = model(image)
    pred_object_list = [str(i)+'_'+x['label'] for i, x in enumerate(pred)]
    return gr.Dropdown.update(choices = pred_object_list, interactive = True)


    
def get_seg(image, model_choice):
    image = img_resize(image)
    model = models[model_choice]
    segment = pipeline("image-segmentation", model=f"{model}")
    pred = segment(image)
    pred_object_list = [str(i)+'_'+x['label'] for i, x in enumerate(pred)]
    seg_box=[]
    for i in range(len(pred)):
        #object_number = int(object.split('_')[0])
        mask_array = np.asarray(pred[i]['mask'])/255
        image_array = np.asarray(image)
    
        mask_array_three_channel = np.zeros_like(image_array)
        mask_array_three_channel[:,:,0] = mask_array
        mask_array_three_channel[:,:,1] = mask_array
        mask_array_three_channel[:,:,2] = mask_array
    
        segmented_image = image_array*mask_array_three_channel
        seg_out=segmented_image.astype(np.uint8)
        seg_box.append(seg_out)  

    return(seg_box,gr.Dropdown.update(choices = pred_object_list, interactive = True))




app = gr.Blocks()


with app:
    gr.Markdown(
            """
            ## Image Dissector 
            
            """)
    with gr.Row():
        with gr.Column():
            image_input = gr.Image(label="Input Image",type="pil")
            model_name = gr.Dropdown(show_label=False, choices=[m for m in models], type="index", value=current_model, interactive=True)
        with gr.Column():
            gal1=gr.Gallery(type="filepath").style(grid=6)
    with gr.Row():
        with gr.Column():
                object_output = gr.Dropdown(label="Objects")

    
    image_input.change(get_seg, inputs=[image_input, model_name], outputs=[gal1,object_output])



app.launch()