Want segmentation model with 25 fps for jetson nano

[fanzh]

if using fakesink, please use the following code:

    if is_aarch64():
        #nvsegvisual.link(transform)
        #transform.link(sink)
        nvsegvisual.link(sink)

[Keshav.A]

Thank you it worked. Max i get is 11 fps with one stream.
I check sudo tegrastats
output is
test1.txt (17.5 KB)
I found one of the CPU core is 100% used. But others are not used to full there might be the scope of increasing the fps i guess what are your view onit.

[fanzh]

thanks for the sharing! can you use “top -H -p ” or jtop check which task cost too much CPU resource?

[fanzh]

There is no update from you for a period, assuming this is not an issue anymore.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

you can add more queue plugin in pipeline. queue is Gstreamer opensource plugin. AYK, there are multiple threads in queue, multiple threads will use more CPU cores.