DetectionTeamUCAS / FPN_Tensorflow

This is a tensorflow re-implementation of Feature Pyramid Networks for Object Detection.
https://github.com/DetectionTeamUCAS/FPN_Tensorflow
MIT License
347 stars 132 forks source link

Training on custom dataset #125

Open Janzeero-PhD opened 4 years ago

Janzeero-PhD commented 4 years ago

Hello, sorry for the probably stupid question. I have seen an instruction in this repo. Ok, in lable_dict.py I see different class names. If I have my own dataset with 1 class, should I remove all this data to leave only my custom class? What dataset = 'pascal' means in this case? Style of annotations? Should I change these notes if I use own custom dataset but labelled in Pascal VOC style? The same questions for changing cfgs.pyand other.