Visualizer().render fails for 16-bit images

[piotrgrubicki]

1. Train a model with 16-bit images using Geti.
2. Export model
3. Use the following piece of code to run inference using Model API
4. The code throws an exception as shown below (ignore the line numbers).

    model = Model.create_model(str(MODEL_PATH))
    image_bgr = cv2.imread(str(IMAGE_PATH), cv2.IMREAD_UNCHANGED)
    if image_bgr is None:
        raise RuntimeError(f"Failed to decode image: {IMAGE_PATH}")
    result = model(image_bgr)
    output = Visualizer().render(image_bgr, result)
    cv2.imwrite(str(OUTPUT_PATH), output)

Traceback (most recent call last):
  File "/home/osboxes/deployment/16bit/demo.py", line 51, in <module>
    main()
    ~~~~^^
  File "/home/osboxes/deployment/16bit/demo.py", line 45, in main
    output = Visualizer().render(image_bgr, result)
  File "/home/osboxes/deployment/16bit/.venv/lib/python3.13/site-packages/model_api/visualizer/visualizer.py", line 85, in render
    image = Image.fromarray(image)
  File "/home/osboxes/deployment/16bit/.venv/lib/python3.13/site-packages/PIL/Image.py", line 3431, in fromarray
    raise TypeError(msg) from e
TypeError: Cannot handle this data type: (1, 1, 3), <u2