Jetson TX2上跑Yolo v3以及fasterRCNN

如何在DL中使用tensorRT（The most-optimized Jetson library is TensorRT）？可以参考下面的文章：
Guide to deploying deep-learning inference networks and deep vision primitives with TensorRT and NVIDIA Jetson.
https://github.com/dusty-nv/jetson-inference

结合DeepStream和Yolo的案例(在nv官网的讨论)：
https://github.com/vat-nvidia/deepstream-plugins

在tx2上移植fasterRCNN:
https://jkjung-avt.github.io/faster-rcnn/
https://jkjung-avt.github.io/making-frcn-faster/

https://devtalk.nvidia.com/default/topic/1028409/jetson-tx2/how-can-i-finetune-the-tensorrt-faster-rcnn-sample-/