simple person detection with Yolov8.

.gitignore

@@ -0,0 +1,2 @@
+.idea
+*.pt

app.py

@@ -0,0 +1,38 @@
+import cv2
+import gradio as gr
+from ultralytics import YOLO
+
+def inference(path:str, threshold:float=0.6):
+    print("trying inference with path", path)
+    if path is None:
+        return None,0
+    model = YOLO('yolov8m.pt')
+
+    model.classes = [0]  # only considering class 'person' and not the 79 other classes...
+    image = cv2.imread(path)
+    outputs = model.predict(source=path, return_outputs=True)
+    for output in outputs:  # mono item batch
+        detections = output['det']
+        counter=0
+        for detection in detections:
+            if detection[4]<threshold:
+                break
+            cv2.rectangle(
+                image,
+                (int(detection[0]), int(detection[1])),
+                (int(detection[2]), int(detection[3])),
+                color=(0, 0, 255),
+                thickness=2,
+            )
+            counter+=1
+    return cv2.cvtColor(image, cv2.COLOR_BGR2RGB), counter
+
+gr.Interface(
+    fn = inference,
+    inputs = [ gr.components.Image(type="filepath", label="Input"), gr.Slider(minimum=0.5, maximum=0.9, step=0.05, value=0.7, label="Confidence threshold") ],
+    outputs = [ gr.components.Image(type="numpy", label="Output"), gr.Label(label="nb of persons detected for given confidence threshold") ],
+    title="Person detection with YOLO v8",
+    description="Person detection, you can tweak the corresponding confidence threshold. Good results even when face not visible.",
+    examples=[ ['data/businessmen-612.jpg'], ['data/businessmen-back.jpg']],
+    allow_flagging="never"
+).launch(debug=True, enable_queue=True)

requirements.txt

@@ -0,0 +1,6 @@
+torch
+torchvision
+ultralytics
+gradio
+opencv-python
+numpy