• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 31 subscribers
  • Views 154 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AWRL6844EVM: Third row Zone Definition

Chun-Ju Wu
Prodigy 80 points
Part Number: AWRL6844EVM

Tool/software:

Hi,

The SBR Users Guide mentions that there are 8 supported zones, should they be zones 0 to 7?

When we tested setting up the third row of space, zones 5, 6, and 7, we always got the following error message:

ClassifierProc DPU Configuration return error:-8003

Does the radar_toolbox_3_10_00_05 version not support the third row area detection yet?

Best

ED

  • +1
    TI__Genius 13126 points

    Hey Chun-Ju,

    Thanks for reaching out regarding the SBR demo. I have looped in our SBR expert to comment further, but it does appear the demo is running amount of memory with the third row defined as indicated by error -8003. You can try to reduce the maximum number of points per zone (default for SBR is 200) or the number of frames to be processed before making a decision (default for SBR is 20) by changing the first and second parameter of featExtrCfg respectively.

    I would also like to note that if you would like to work towards three row coverage, it will require additional retuning of the point cloud via CFAR and other front-end adjustments and model retraining since the current model is built primarily for two row coverage. 

    Regards,

    Kristien

  • 0 in reply to Kristien Everett
    Prodigy 80 points

    Hi Kristien,

    The documentation indicates that the parameter featExtrCfg should be modified as little as possible.

    Considering your description as well, we will have to wait until the model training method is released before we can proceed with the third row settings.

    I heard that the training method for ML models will be released in June. Is there an estimated time now?

    Thank you for your answer.

    Regards,

    ED