Advice needed: Classifying Object Detection Bounding Boxes

@aurel could I ask also to have the time extended for my other project, “RSPB_MLCAM_OD_PROTOTYPE_SOMETHINGVSNOTHING” please?

I’m taking @dansitu’s advice and attempting to recognise something vs nothing, and it seems to be a lot more successful than the Mammal vs Bird vs Bird in Flight model I was attempting to train.

Also, is it the case that the more data I feed into it the longer it takes to train the model? It didn’t have trouble with the initial amount of images I uploaded, about 750, but as I add more it seems to start struggling a little. If so, I’ll keep that in mind and maybe tailor my images a bit better.

Thanks in advance for any help/info you can offer!

@TechDevTom done!

Yes it will take longer the more images you add.

Aurelien

1 Like

Cheers @aurel!

Thought that would most likely be the case, will keep that in mind.

2 Likes