Ressource to get started converting TensorRT python inference to C++ Inference

Hello everyone,
I am a pretty advanced C++ and python developer, however I am very new to AI programming. I’ve been toying a round a small first project to test and decided to try and convert TRT_POSE ( from python inference to C++ inference. I’ve written a small python equivalent of the live_demo.ipynb file to run with opencv on a ubuntu 18.04 machine. Now, I would like to convert the inference script to C++ on Windows 10. I’ve been able to compile all the example that come with TensorRT like sampleMNIST or sampleOnnxMNIST and I’ve successfully converted the .pth files to .onnx on linux. Now I would be ready to take my learning further. Any ressources/books or cues on how to get started? Thanks a lot!