Problem on deploying LPD model in python

Hi, I am deploying LPD model in python but the results I got are wrong. The document says the output is four normalized bounding-box parameters (xc, yc, w, h) and confidence value per output class. I think the bounding-box parameters shouldn’t be bigger than 1 but most of my results are bigger than 1. I think perhaps my inputs are wrong or misunderstand something.

I checked my inputs and the input images are satisified:

  • Color Images of resolution 640 X 480 X 3 (W x H x C)(RGB)
  • Channel Ordering of the Input: NCHW, where N = Batch Size, C = number of channels (3), H = Height of images (480), W = Width of the images (640)
  • Input scale: 1/255.0

The model is usa_pruned.etlt downed from NGC and convertered to usa_pruned_int8.trt.

Can detectnet_v2 infer or deepstream works well when you deploy the LPD model?

I didn’t try that. I directly deploy the LPD model in python.

Refer to Run PeopleNet with tensorrt - #21 by carlos.alvarez

1 Like

It solved my problem! Thank you!