This pull request adds support for the YOLOv11 object detection model to the lite TensorRT-based computer vision toolkit. The main changes include implementing the TRTYOLOV11 class for inference, updating the build system, and adding a corresponding test example. These updates allow users to run YOLOv11 inference with proper preprocessing, postprocessing, and NMS, similar to other YOLO models in the toolkit.

YOLOv11 Model Integration

• Added implementation of the TRTYOLOV11 class in trt_yolov11.cpp and its header, including preprocessing (letterbox resizing, normalization), inference, and postprocessing (bounding box decoding and NMS). [1] [2]
• Declared the TRTYOLOV11 class in the core header and registered it in the type system for easy use as lite::trt::cv::detection::YOLOV11. [1] [2] [3] [4]

Build System and Example

• Updated CMake configuration to add a new executable for YOLOv11 (lite_yolov11).
• Added a new test example test_lite_yolov11.cpp to demonstrate and validate YOLOv11 inference using TensorRT.