File size: 3,346 Bytes
3c72bd0
 
 
 
 
 
a897877
3c72bd0
 
 
 
 
8a42819
a897877
3c72bd0
 
 
 
8a42819
 
3c72bd0
 
 
 
 
 
8a42819
 
3c72bd0
8a42819
3c72bd0
 
8a42819
 
3c72bd0
 
 
 
 
 
 
8a42819
3c72bd0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
import os
import gradio as gr
from transformers import pipeline, DetrForObjectDetection, DetrConfig, DetrImageProcessor
import numpy as np
import cv2
from PIL import Image

def load_model(model_name, threshold):
    config = DetrConfig.from_pretrained(model_name, threshold=threshold)
    model = DetrForObjectDetection.from_pretrained(model_name, config=config)
    image_processor = DetrImageProcessor.from_pretrained(model_name)
    return pipeline(task='object-detection', model=model, image_processor=image_processor)
od_pipe = load_model("facebook/detr-resnet-101", 0.25)  # Default model

def draw_detections(image, detections, model_name):
    np_image = np.array(image)
    np_image = cv2.cvtColor(np_image, cv2.COLOR_RGB2BGR)
    for detection in detections:
        if 'mask' in detection:
            # Interpret and visualize segmentation mask
            mask = detection['mask']
            color = np.random.randint(0, 255, size=3)
            mask = np.round(mask * 255).astype(np.uint8)
            mask = cv2.resize(mask, (image.width, image.height))
            mask_image = np.stack([mask]*3, axis=-1)
            np_image[mask == 255] = np_image[mask == 255] * 0.5 + color * 0.5
        if 'box' in detection:
            # Visualize bounding box
            box = detection['box']
            x_min, y_min, x_max, y_max = [int(b) for b in [box['xmin'], box['ymin'], box['xmax'], box['ymax']]]
            cv2.rectangle(np_image, (x_min, y_min), (x_max, y_max), (0, 255, 0), 2)
    final_image = cv2.cvtColor(np_image, cv2.COLOR_BGR2RGB)
    return Image.fromarray(final_image)

def get_pipeline_prediction(model_name, threshold, pil_image):
    global od_pipe
    od_pipe = load_model(model_name, threshold)
    try:
        if not isinstance(pil_image, Image.Image):
            pil_image = Image.fromarray(np.array(pil_image).astype('uint8'), 'RGB')
        result = od_pipe(pil_image)
        print("Detection Output:", result)  # Debug: Check the output structure
        processed_image = draw_detections(pil_image, result, model_name)
        description = f'Model used: {model_name}, Detection Threshold: {threshold}'
        return processed_image, result, description
    except Exception as e:
        return pil_image, {"error": str(e)}, "Failed to process image"
with gr.Blocks() as demo:
    with gr.Row():
        with gr.Column():
            gr.Markdown("## Object Detection")
            inp_image = gr.Image(label="Upload your image here")
            model_dropdown = gr.Dropdown(choices=["facebook/detr-resnet-50", "facebook/detr-resnet-50-panoptic", "facebook/detr-resnet-101", "facebook/detr-resnet-101-panoptic"], value="facebook/detr-resnet-101", label="Select Model")
            threshold_slider = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, value=0.25, label="Detection Threshold")
            run_button = gr.Button("Detect Objects")
        with gr.Column():
            with gr.Tab("Annotated Image"):
                output_image = gr.Image()
            with gr.Tab("Detection Results"):
                output_data = gr.JSON()
            with gr.Tab("Description"):
                description_output = gr.Textbox()
    run_button.click(get_pipeline_prediction, inputs=[model_dropdown, threshold_slider, inp_image], outputs=[output_image, output_data, description_output])
demo.launch()