Drastic Drop in the Frame Rate on the Jetson Nano

[contactnikhilrb]

I’ve implemented a facial detection + recognition code on the Jetson Nano, the facial detection part is implemented using MediaPipe 0.8.5 (BlazeFace model) and runs on the GPU of the Jetson nano, while the recognition is implemented using DLib’s face_recognition library and runs on the CPU of the Jetson Nano.

Commenting out the facial recognition part of the code, the facial detection part of the code runs at about 30 fps. But, when I try running both the detection+recognition the overall frame rate drops to about 5 fps.
Need help with how I can avoid this drastic drop in the overall frame rate on the Jetson Nano.

[AastaLLL]

Hi,

Did you use VGG for recognition?
Usually, recognition requires much more computational resources to complete.

Please check with tegrastats if the GPU is fully utilized.
If yes, the fps might be limited due to the Nano GPU resources.

$ sudo tegrastats 

Thanks.

[contactnikhilrb]

@AastaLLL

Did you use VGG for recognition?

No.
Can you please share alternative to DLib’s face_recognition library that can be used on the Jetson nano.

[AastaLLL]

Hi,

Due to data privacy issues, we don’t have a face recognition sample.
But you should be able to find some by googling.

Thanks.

[contactnikhilrb]

Alright, thanks