Spaces:
Sleeping
Sleeping
File size: 4,073 Bytes
40ceb45 |
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 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 |
import random
import requests
from PIL import Image, ImageDraw, ImageFont
import gradio as gr
import torch
from transformers import pipeline
# Load object detection pipeline
finder = pipeline(
"object-detection",
model="facebook/detr-resnet-50",
torch_dtype=torch.bfloat16
)
# Generate a random bright color
def random_color():
return tuple(random.randint(50, 255) for _ in range(3))
# Main detection function
def Searchy(input_type, url, file, threshold):
# Load image based on choice
if input_type == "Image URL":
image = Image.open(requests.get(url, stream=True).raw).convert("RGB")
else:
image = Image.open(file).convert("RGB")
# Run object detection
output = finder(image, return_tensors="pt")
# Draw bounding boxes
draw = ImageDraw.Draw(image)
for item in output:
if item['score'] < threshold:
continue
label = item['label']
score = item['score']
box = item['box']
xmin, ymin, xmax, ymax = box['xmin'], box['ymin'], box['xmax'], box['ymax']
# Random bright color for each object
color = random_color()
# Draw rectangle
draw.rectangle([(xmin, ymin), (xmax, ymax)], outline=color, width=1)
# Dynamic font size
font_size = max(5, int(image.width * 0.025))
try:
font = ImageFont.truetype("arial.ttf", font_size)
except:
font = ImageFont.load_default()
# Label text
text = f"{label} {score:.2f}"
# Measure text size
bbox = draw.textbbox((0, 0), text, font=font)
text_width = bbox[2] - bbox[0]
text_height = bbox[3] - bbox[1]
# Background rectangle for text
draw.rectangle(
[(xmin, ymin - text_height - 6), (xmin + text_width + 6, ymin)],
fill=color
)
# Text in contrasting color
draw.text((xmin + 3, ymin - text_height - 3), text, fill="black", font=font)
return image
# Toggle visibility of inputs
def toggle_inputs(choice):
if choice == "Image URL":
return gr.update(visible=True), gr.update(visible=False)
else:
return gr.update(visible=False), gr.update(visible=True)
# Theme
theme = gr.themes.Soft(
primary_hue="orange",
secondary_hue="blue",
neutral_hue="slate"
).set(
button_primary_background_fill="linear-gradient(90deg, #ff7e5f, #feb47b)",
button_primary_text_color="white",
block_background_fill="#f9f9f9",
block_border_color="#ddd"
)
# Build Gradio UI
with gr.Blocks(theme=theme, css=open("style.css").read()) as demo:
gr.Markdown("## 🖼️ Object Finder — Choose URL or Upload", elem_id="header")
with gr.Row():
with gr.Column(scale=1, elem_id="input_panel"):
input_type = gr.Radio(
["Image URL", "Upload Image"],
value="Image URL",
label="Select Input Type"
)
url_input = gr.Textbox(
label="Enter Image URL",
placeholder="https://example.com/image.jpg",
visible=True
)
file_input = gr.File(
label="Upload Image",
type="filepath",
visible=False
)
threshold_slider = gr.Slider(
minimum=0.0, maximum=1.0, value=0.5, step=0.01,
label="Confidence Threshold"
)
submit_btn = gr.Button("🔍 Detect Objects", elem_id="detect_btn")
with gr.Column(scale=2):
output_img = gr.Image(
type="pil",
label="Image with Detected Objects",
image_mode="RGB",
height=500,
elem_id="output_image",
show_download_button=True
)
# Events
input_type.change(toggle_inputs, inputs=input_type, outputs=[url_input, file_input])
submit_btn.click(Searchy, inputs=[input_type, url_input, file_input, threshold_slider], outputs=output_img)
demo.launch()
|