Low "interval" causes loss frame

Accelerated Computing Intelligent Video Analytics DeepStream SDK
#1

• Hardware Platform (Jetson / GPU) GPU
• DeepStream Version 6.1
• TensorRT Version 8.4.2.4
• NVIDIA GPU Driver Version 510.47.03
• Issue Type questions or bugs

I set pgie to “interval=0”, I detect more frame outputs.
I set pgie to “interval=5”, I detect less frame outputs.

However, “interval=5” mode has some frame outputs, “interval=0” can’t detect that. It means “interval=0” will cause loss frames?

I already checked my computer gpu usage and cpu usage and memory is all ok. What causes that?

My code is from deepstream-appsrc-test.
This reason is from appsrc pipeline? How to fix that?

#3

Do you mean some objects are missed with “interval=0”?

#5

What are these?

#6

These is example.

Or

==========================

“interval=5” Output =>
car1
car1
car2
car2
car3
car3
car4
car4

“interval=0” Output =>

car1
car1
car1
car1
car2
car2
car2
car2
car4
car4
car4
car4

=======================

#7

Actually there is only confidence print out in your code. I’m confused with your description. Can you send the correct code?

#8

Okay, I already sent a message to you.

#9

I think maybe the speed of my computer can’t support high framerate.

deepstream-app : drop-frame-interval <== do you know how do I add to “deepstream-appsrc-test”?

Because I know “interval” doesn’t match my need. it still run full fps, just inferece a part of those.