Darknet on Jetson Nano with Intel RealSense D435i

Hi, i want to ask something, im currently working on a project to run yolov4-tiny using darknet on Nvidia Jetson Nano using Intel RealSense D435i, im have successfully run the darknet using Intel RealSense D435i, but the problem is the video resolution is 1920 x 1080, which is really high and makes the fps really low, im wondering if there is a way to run darknet but lower the resolution, for example to 480p maybe.

The command i use to run darknet is :
./darknet detector demo cfg/coco.data cfg/yolov4-tiny.cfg yolov4-tiny.weights -c 0

Any suggestion would be really appreciated,
Thank You

Hi,

In general, the library will downscale the input image to the model input size.
Changing the camera resolution might not change the real inference complexity.

Could you check the GPU loading when running the darknet first?

$ sudo tegrastats

Thanks.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.