DCF Filter Learning doesn't seem to work as expected


• Hardware Platform: Jetson Xavier NX
• DeepStream Version: 6.1.1
• JetPack Version: 5.0.2

I’m trying to tweak nvtracker according to the DeepStream Developer Guide, mainly filterLr, filterChannelWeightsLr and gaussianSigma.

My understanding is that these parameters control the rate/measure of a process which involves increasing the confidence over time: “If the visual appearance of the target objects is expected to vary quickly over time, one may employ a high learning rate for better adaptation of the correlation filter to the changing appearance”.

No matter how much I tweak these parameters, the confidence is getting lower when the object is changing “away” from the initial appearance, and higher when it’s changing “towards” the initial appearance (which makes sense). But I’m not seeing any improvement in confidence when an object is static in visual appearance, even over extended periods.

Any help will be much appreciated.

Can you share how you check it? Can you share the test video? So we can reproduce and have a check.

Hello @kesong,

Sure, I’ll arrange a test video. I’m simply starting a track on a specific object and add the confidence to OSD, let’s say the confidence starts at ~0.7, then I change the object attitude or rotate it a bit, the confidence is going down to around ~0.4, but when I hold it still (after that attitude change) the confidence never goes up.

In the meantime, can you please tell me if my assumption regarding DCF Filter Learning increasing confidence over time is right?

Thank you.

Hi @kesong, can you please verify my claim regarding DCF Filter Learning’s behavior?

It is yes in theory. But I need your video to check the details.