JETSON NANO -- YOLOv3 Real Time Object Detection With Camera

Hello there,

I’m new in research and application of Computer Vision Application and it’s the first time I work on Jetson nano platform. I want to design traffic sign detector with Jetson nano using YOLO algorithms. What adjustments should I make, help me where should I start and how should I proceed ?

I want to do something as shown in the attached picture.



Suppose you already have a dataset for your use case.
Try following two steps to deploy a customized model on Nano.

1. Train a customized YOLO model with our TLT toolkit:

2. Deploy the model with Deepstream for camera pipeline: