adjust visualisation style, allow draw_bbox() to show person id

Add method to free buffers

Implement pycuda backend for inference with TensorRT engine

Add 'get_model_format_and_input_shape' helper function

Return and show bounding box confidence

Deal with multiple cameras

Show bounding box on screen too

Internalize input frame buffer

Unify batched and non-batched versions

Support Inference over batch with TensorRT Engine Model

Use ONNXruntime instead of ONNX.checker.check_model to detect ONNX model

Resize in keeping aspect ratio in visualization

fix postprocess_batch method in RTMO_GPU_Batch class

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

Add support TensorRT engine support for RTMO

Load ./libmmdeploy_tensorrt_ops.so for TensorRT EP if available

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

Make ONNX models compatible with ONNXruntime's TensorrtExecutionProvider

Add new class 'RTMO_GPU_Batch' that can perform inference on batch of images

Optimization for CUDAExecutionProvider

RTMO ONNX Models from MMPOSE range from Tiny to Large