File size: 4,565 Bytes
ec41162 fa8b60d ec41162 676de75 ec41162 |
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 |
import os
import logging
import gradio as gr
from PIL import Image as PILImg
from iteach_toolkit.DHYOLO import DHYOLODetector
# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
def detect_objects(selected_model, input_image, conf, imu_threshold, detections):
try:
model_path = model_options[selected_model]
dhyolo = DHYOLODetector(model_path)
# Save the input image temporarily to a path for processing
input_image_path = "dhyolo_temp_input_image.jpg"
input_image.save(input_image_path)
# Perform prediction on the image
orig_image, detections = dhyolo.predict(input_image_path, conf, imu_threshold, detections)
# Log the detections
logger.info("Detections: %s", detections)
# Plot the bounding boxes on the original image
orig_image, image_with_bboxes = dhyolo.plot_bboxes(attach_watermark=True)
# Convert the image (with bounding boxes) from a NumPy array to a PIL Image for display.
pil_img_with_bboxes = PILImg.fromarray(image_with_bboxes)
return input_image, pil_img_with_bboxes
except FileNotFoundError as e:
logger.error("File not found: %s", e)
return None, None
except Exception as e:
logger.error("An error occurred: %s", e)
return None, None
def load_test_images():
"""Load images from the test_imgs directory."""
test_imgs_dir = os.path.join(os.getcwd(), "test_imgs")
logger.info("Loading images from: %s", test_imgs_dir) # Log the directory path
return [f for f in os.listdir(test_imgs_dir) if f.lower().endswith(('.png', '.jpg', '.jpeg'))]
# Gradio interface
def create_interface():
with gr.Blocks() as demo:
# Center the title using HTML
gr.Markdown("<h1 style='text-align: center;'>🚪🔍 DHYOLO DoorHandle Object Detection</h1>")
# Add paper description with correct hyperlink syntax
gr.Markdown("<h2 style='text-align: center;'>📚 iTeach: Interactive Teaching for Robot Perception using Mixed Reality</h2>")
# Add the project link using HTML <a> tag
gr.Markdown("<h2 style='text-align: center;'>🌐 Project Link: <a href='https://irvlutd.github.io/iTeach/' target='_blank'>iTeach Project</a></h2>")
with gr.Row():
with gr.Column():
# Load a default image for input
default_image_path = os.path.join(os.getcwd(), "test_imgs", "jpad-irvl-test.jpg") # Update with the correct default image name
# Attempt to open the default image and log if it fails
try:
default_image = PILImg.open(default_image_path)
except FileNotFoundError:
logger.error("Default image not found at: %s", default_image_path)
default_image = None # Set to None or a placeholder image
input_image = gr.Image(type="pil", label="Input Image", value=default_image)
# Pretrained model paths
cwd = os.getcwd()
global model_options
model_options = {
"dh-yolo-v1-pb-ddf-524": f'{cwd}/pretrained_ckpts/dh-yolo-v1-pb-ddf-524.pt',
"dh-yolo-exp27-pb-1008": f'{cwd}/pretrained_ckpts/dh-yolo-exp27-pb-1008.pt',
"dh-yolo-exp31-pb-1532": f'{cwd}/pretrained_ckpts/dh-yolo-exp-31-pb-1532.pt',
"dh-yolo-exp31-pl-1532": f'{cwd}/pretrained_ckpts/dh-yolo-exp-31-pl-1532.pt'
}
model_path = gr.Dropdown(choices=list(model_options.keys()), label="Select Pretrained Model", value="dh-yolo-v1-pb-ddf-524")
conf = gr.Slider(label="Confidence Threshold", minimum=0.0, maximum=1.0, step=0.01, value=0.5)
imu_threshold = gr.Slider(label="IoU Threshold", minimum=0.0, maximum=1.0, step=0.01, value=0.5)
detections = gr.Slider(label="Max number of Detections", minimum=1, maximum=100, step=1, value=10)
with gr.Column():
output_image = gr.Image(label="Output Image with DH-YOLO Detections", type="pil")
detect_button = gr.Button("Run")
# Detect button functionality
detect_button.click(detect_objects, inputs=[model_path, input_image, conf, imu_threshold, detections],
outputs=[input_image, output_image])
return demo
if __name__ == "__main__":
demo = create_interface()
demo.launch()
|