video_object_detection.py issue

[Laurent Weingart]

Hello,

I simply cloned the repo and tried to run the video_objet_detection.py, but it directly fails with:

$ python3 video_object_detection.py
Traceback (most recent call last):
  File "video_object_detection.py", line 4, in <module>
    from yolov8 import YOLOv8
  File "/home/jetson/git/ONNX-YOLOv8-Object-Detection/yolov8/__init__.py", line 1, in <module>
    from .YOLOv8 import YOLOv8
  File "/home/jetson/git/ONNX-YOLOv8-Object-Detection/yolov8/YOLOv8.py", line 6, in <module>
    from yolov8.utils import xywh2xyxy, draw_detections, multiclass_nms
  File "/home/jetson/git/ONNX-YOLOv8-Object-Detection/yolov8/utils.py", line 108, in <module>
    def draw_box( image: np.ndarray, box: np.ndarray, color: tuple[int, int, int] = (0, 0, 255),
TypeError: 'type' object is not subscriptable

I had to remove the type description to make it work (i.e. color: tuple[int, int, int] = (0, 0, 255) turned into color=(0, 0, 255)

Maybe there is a way to fix this that I'm not aware of yet ?

Cheers

[Michael Lehning]

For older python version like 3.8 use the following modifications for utils.py:

1. Add from typing import Tuple in the file header
2. Replace tuple with Tuple in draw_box and draw_text.

Which Python version do you use?