• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 32 subscribers
  • Views 162 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.

IWRL6432AOPEVM: Combine Minor Motion along with Human Classifier

Sathishraj Pushpanathan
Prodigy 20 points
Part Number: IWRL6432AOPEVM

Tool/software:

I used the demo, Motion and Presence Detection. In that Configuration,

I need Ceiling Mount, to capture the human location data. I need to enable the human/non-human classifier and also i need minor motion of human.

"sensorPosition 0 0 2.5 0 90 \n \r",

"minorStateCfg 5 4 40 2 1 40 8 20 \n \r",

"clusterCfg 1 0.5 2 \n \r",

"gatingParam 3 2 2 2 4 \n\r",

"stateParam 3 3 12 50 5 200 \n\r",

"allocationParam 6 10 0.1 4 0.5 20 \n\r",

"maxAcceleration 0.4 0.4 0.1 \n\r",

"trackingCfg 1 2 100 5 61.3 191.7 100 \n\r",

"presenceBoundaryBox -2 2 -2 2 0 2  \n\r",

"microDopplerCfg 1 0 0.5 0 1 1 12.5 87.5 1 \n\r",

"classifierCfg 1 3 4 \n\r",

When I put the above configuration, code is not executing. In the tuning guide it is mentioned that we can't use both trackingCfg and minorStateCfg at the same time.

If only Minor motion is enabled, in ceiling mount, sensor is picking up the small curtain movement also as human minor motion. It is a false detection. So, If I get minor motion along with, whether it is a Human or Non-Human, this will solve the problem. But I can't enable both!!! How to solve this?

In human_non-human_classification, it is using a model trained for human at wall mount. How can use that model or how to create a model that classify human in ceiling mount.

  • 0
    TI__Genius 11750 points

    Hi

    Thanks for your query. Please allow us a couple of days to respond

    Regards

  • 0 in reply to Abhishek Karkisaval
    TI__Genius 11750 points

    Hi Satishraj,

    The current classifier is trained for wall mount scenario and it will not work for the ceiling mount case. Model will have to retrained to make it effective for the ceiling mount case. Classifier is still slightly exploratory and may not provide the accuracy that might be fit for a production scenario.

    It is recommended that ceiling mount tracking itself is tuned to track human movements correctly and reject some of the minor movements like curtain movement 

    Regards