Spaces:
Running
on
Zero
Running
on
Zero
Update models/detection/detector.py
Browse files- models/detection/detector.py +86 -73
models/detection/detector.py
CHANGED
@@ -1,73 +1,86 @@
|
|
1 |
-
import os
|
2 |
-
import numpy as np
|
3 |
-
from PIL import Image, ImageDraw
|
4 |
-
import logging
|
5 |
-
from ultralytics import YOLO
|
6 |
-
from utils.model_downloader import download_model_if_needed
|
7 |
-
|
8 |
-
logger = logging.getLogger(__name__)
|
9 |
-
|
10 |
-
class ObjectDetector:
|
11 |
-
"""
|
12 |
-
Generalized Object Detection Wrapper for YOLOv5, YOLOv8, and future variants.
|
13 |
-
"""
|
14 |
-
|
15 |
-
def __init__(self, model_key="yolov5n-seg",
|
16 |
-
"""
|
17 |
-
Initialize the Object Detection model.
|
18 |
-
|
19 |
-
Args:
|
20 |
-
model_key (str): Model identifier as defined in model_downloader.py.
|
21 |
-
weights_dir (str): Directory to store/download model weights.
|
22 |
-
device (str): Inference device ("cpu" or "cuda").
|
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 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import numpy as np
|
3 |
+
from PIL import Image, ImageDraw
|
4 |
+
import logging
|
5 |
+
from ultralytics import YOLO
|
6 |
+
from utils.model_downloader import download_model_if_needed
|
7 |
+
|
8 |
+
logger = logging.getLogger(__name__)
|
9 |
+
|
10 |
+
class ObjectDetector:
|
11 |
+
"""
|
12 |
+
Generalized Object Detection Wrapper for YOLOv5, YOLOv8, and future variants.
|
13 |
+
"""
|
14 |
+
|
15 |
+
def __init__(self, model_key="yolov5n-seg", device="cpu"):
|
16 |
+
"""
|
17 |
+
Initialize the Object Detection model.
|
18 |
+
|
19 |
+
Args:
|
20 |
+
model_key (str): Model identifier as defined in model_downloader.py.
|
21 |
+
weights_dir (str): Directory to store/download model weights.
|
22 |
+
device (str): Inference device ("cpu" or "cuda").
|
23 |
+
"""
|
24 |
+
repo_map = {
|
25 |
+
"yolov5n": ("ultralytics/yolov5", "yolov5n.pt"),
|
26 |
+
"yolov8n": ("ultralytics/yolov8", "yolov8n.pt"),
|
27 |
+
# Add more if needed
|
28 |
+
}
|
29 |
+
|
30 |
+
if model_key not in repo_map:
|
31 |
+
raise ValueError(f"Unsupported model_key: {model_key}")
|
32 |
+
|
33 |
+
repo_id, filename = repo_map[model_key]
|
34 |
+
|
35 |
+
weights_path = hf_hub_download(
|
36 |
+
repo_id=repo_id,
|
37 |
+
filename=filename,
|
38 |
+
cache_dir="models/detection/weights"
|
39 |
+
)
|
40 |
+
|
41 |
+
self.device = device
|
42 |
+
self.model = YOLO(weights_path)
|
43 |
+
|
44 |
+
def predict(self, image: Image.Image):
|
45 |
+
"""
|
46 |
+
Run object detection.
|
47 |
+
|
48 |
+
Args:
|
49 |
+
image (PIL.Image.Image): Input image.
|
50 |
+
|
51 |
+
Returns:
|
52 |
+
List[Dict]: List of detected objects with class name, confidence, and bbox.
|
53 |
+
"""
|
54 |
+
logger.info("Running object detection")
|
55 |
+
results = self.model(image)
|
56 |
+
detections = []
|
57 |
+
for r in results:
|
58 |
+
for box in r.boxes:
|
59 |
+
detections.append({
|
60 |
+
"class_name": r.names[int(box.cls)],
|
61 |
+
"confidence": float(box.conf),
|
62 |
+
"bbox": box.xyxy[0].tolist()
|
63 |
+
})
|
64 |
+
logger.info(f"Detected {len(detections)} objects")
|
65 |
+
return detections
|
66 |
+
|
67 |
+
def draw(self, image: Image.Image, detections, alpha=0.5):
|
68 |
+
"""
|
69 |
+
Draw bounding boxes on image.
|
70 |
+
|
71 |
+
Args:
|
72 |
+
image (PIL.Image.Image): Input image.
|
73 |
+
detections (List[Dict]): Detection results.
|
74 |
+
alpha (float): Blend strength.
|
75 |
+
|
76 |
+
Returns:
|
77 |
+
PIL.Image.Image: Image with bounding boxes drawn.
|
78 |
+
"""
|
79 |
+
overlay = image.copy()
|
80 |
+
draw = ImageDraw.Draw(overlay)
|
81 |
+
for det in detections:
|
82 |
+
bbox = det["bbox"]
|
83 |
+
label = f'{det["class_name"]} {det["confidence"]:.2f}'
|
84 |
+
draw.rectangle(bbox, outline="red", width=2)
|
85 |
+
draw.text((bbox[0], bbox[1]), label, fill="red")
|
86 |
+
return Image.blend(image, overlay, alpha)
|