sipeed / MaixPy-v1

MicroPython for K210 RISC-V, let's play with edge AI easier
https://wiki.sipeed.com/maixpy
Other
1.68k stars 439 forks source link

How to interprete custom object detection model #487

Open MertenJakobza opened 1 year ago

MertenJakobza commented 1 year ago

Describe the bug I have a custom object detection model that analyses a 256x256 Image. It should output a 32x32x16 Array or someting like that. Because it outputs a 32x32 Heatmap with 15 classes each plus the background class. How do I have to setup the kpu?

kpu.set_outputs(task, 0, 32, 32, 16) I've used this until this point, but I'm not sure if this is right?

And how can I acces the fmap after classification for each of the classes so that I can then produce for example a list of bounding boxes? Or can I turn it into an array so that I can use it that way?

Screenshots image

Please complete the following information