YOLO v8 issue with Coral TPU

@mailseth yeah I'd seen those, but the problem is that it seems like there's a bunch of people trying to figure out how to load custom models in codeproject.ai for using those -- there's even a post on their forum from an hour ago where a team member says "Unfortunately, custom models for Coral are not currently supported."

Would you mind sharing how you were able to accomplish using that pre-release model?
 
  • Like
Reactions: wasserfloh
I'm actually running these tests with the objectdetection_coral_mulitpu.py file on the CLI, independent of most of the rest of CPAI. It's just calling the tpu_runner.py file for my local testing purposes. (I don't even have the CPAI server running on my machine right now.)

So I'm running the pre-release model simply by specifying it as an option to objectdetection_coral_mulitpu.py, but unfortunately it's not going to be a very useful method for anyone else here.
 
  • Like
Reactions: wasserfloh
Possibly. I found it has a CPU-only instruction midway through the model so it doesn’t run as well as you’d think it should. Everything after that instruction ends up also being CPU-only. (Unless you’re running multi-TPUs, because then you can run the remaining instructions on the next TPU segment.)
 
  • Like
Reactions: CanCuba
Mailseth
I tried to load objectdetection_coral_multitpu.py as have two dual coral tpu's and getting a lot of detections of (bed and skies) in my front yard also (traffic lights). Coral tpu wasted picking up those silly items. All tpu's function fine when load muti-tpu using objectdetection_coral_adapter.py with yoloV8 medium time are 10 ms . This is in codeproject 2.9.5 . Reason for trying to load objectdetection_coral_multitpu.py is to run yolov8 ipcam-general from mikelud that only detects person and vehicles. I have run the ipcam-gemeral with objectdetection_coral_adapter.py but it still detects traffic lights ,tv,skie,airplanes. Any help would be great.

I get the below error in codeproject

06:28:21:Started Object Detection (Coral) module
06:28:22:objectdetection_coral_multitpu.py: Using PIL for image manipulation (Either OpenCV or numpy not available for this module)
06:28:22:objectdetection_coral_multitpu.py: usage: objectdetection_coral_multitpu.py [-h] -m MODEL [MODEL ...] -i INPUT
06:28:22:objectdetection_coral_multitpu.py: [-l LABELS] [-t THRESHOLD]
06:28:22:objectdetection_coral_multitpu.py: [-n NUM_TPUS] [-o OUTPUT] [-c COUNT]
06:28:22:objectdetection_coral_multitpu.py: [-d]
06:28:22:objectdetection_coral_multitpu.py: objectdetection_coral_multitpu.py: error: the following arguments are required: -m/--model, -i/--input

Or am I approaching this wrong to being able to run mikeluds yolov8 ipcam-general on the coral tpu?

As new to this so any guidance would be great.
 

Attachments

  • 2 dual corels installed.jpg
    2 dual corels installed.jpg
    8.8 KB · Views: 0
Last edited:
You must log in or register to reply here.
Top Bottom