Sheila Mae Craus

@RangeKing May I use this diagram for academic purposes? I will cite a reference.

@glenn-jocher @AyushExel I think I am also experiencing the same error. I only got one output tensor, instead of 4 tensors. I tried deploying my yolov8s custom model on TensorFlow...

@h-pozuelo is that also what you're figuring out right now?

@h-pozuelo Are you considering using ONNX instead? I think this blog discusses the post-process: Blog: https://docs.openvino.ai/2022.3/notebooks/230-yolov8-optimization-with-output.html#compare-performance-object-detection-models Notebook: https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/230-yolov8-optimization/230-yolov8-optimization.ipynb I have not tried it myself. If you have successfully achieved it...