-1

I would like to train a object detection model with Tensorflow on my machine.But when I use Tensorflow Object Detection API create a model (it works very well), I convert the model to int8 tflite, but it can't run on my OpenMV H7 Plus. So I wonder which step wrong, and how can I finish this task.

my model training process is here training

Sincere blessings to those who help me！

to train a object detection model with Tensorflow on my machine

Hao_Leng
  • 1

1 Answers1

0

Firstly, the AI models contains a lot of "layers" and if you want to run a spesific AI model, you have to run theese layers with your microprocessor. Also, all microprocessors and their firmware can running only spesific layers.


As far as I know, the OpenMV H7 Plus board cannot inference with any object detection model except [FOMO][1] and [MCUNet(TinyEngine)][2]. Because the OpenMV H7 Plus board is compatible with certain layers. Check out this [link][3] for the compatible layers. The non-commented layer names are available for the board. The commented's will be available soon that Mr. Agyeman told so. I can also advice to you [EdgeImpulse][4]'s local training feature.

If you have any questions, I'm here for any help!

BehicMV
  • 11
  • 3
  • As it’s currently written, your answer is unclear. Please [edit] to add additional details that will help others understand how this addresses the question asked. You can find more information on how to write good answers [in the help center](/help/how-to-answer). – Community May 18 '23 at 04:17