Limelight Detector Crashes

I’m on 2023.4 and the image uploads but trying to use the pipeline with the model crashes and limelight restarts. I was using the detector class. Has anyone had success uploading a model?

Here’s the model if you get a chance to try it: [FRC2023/limelight/converted_edgetpu at main · team6637/FRC2023 · GitHub]
(FRC2023/limelight/converted_edgetpu at main · team6637/FRC2023 · GitHub)

We are on a 2+

@Brandon_Hjelstrom (thank you)

1 Like

This has been resolved. We trained a classifier on Teachable Machine and were uploading it to the detecter option. Thanks Brandon!

1 Like

Wow! @Brandon_Hjelstrom thank you for the support you provided and the amazing product. We have a limelight 2+ running reflective tape detection, April tag detection and a Neural Network detector. I’m so impressed and it’s an approachable product for students. Can’t say it enough, thank you so much!

3 Likes

@ericjanofski Thanks for taking the time to detail all of the issues you experienced. The next update has a few changes based directly on your feedback to make the new features even easier to use.

1 Like

[Inserts McMahon emoji]

Ha, thanks man, your support is unparalleled:)

We are having an issue with our limelight 2+ where we trained a classifier on Teachable Machine and uploaded to the detector option. Now the detector option freezes and the limelight lights blink and that is it.
Are you saying that your issues was that you trained a classifier on Teachable Machine and uploaded it to the detector option was wrong? If so, where/how do you train the detector option?

Thanks,

Todd

Hey there! Yes, the issue was we trained a classifier and uploaded it to the detector, which of course is incorrect. My appologies, I"m not sure how to resolve your issue, it’s been a year since I was in there and my brain memory isn’t that persistent. I also need a ram upgrade :slight_smile:

lol! Thanks for responding! I could use a ram upgrade as well.