Return and show bounding box confidence

Show bounding box on screen too

Internalize input frame buffer

Unify batched and non-batched versions

Show FPS on demo video

Resize in keeping aspect ratio in visualization

Add model format (ONNX/Engine) & input size detection (based on file header, not on filename) for RTMO

Add support TensorRT engine support for RTMO

Add option --yolo_nas_pose, used to read YOLO NAS Pose model instead of RTMO

Optimization for CUDAExecutionProvider

RTMO ONNX Models from MMPOSE range from Tiny to Large