search

IDEA-Research / MaskDINO

[CVPR 2023] Official implementation of the paper "Mask DINO: Towards A Unified Transformer-based Framework for Object Detection and Segmentation"
Apache License 2.0
1.15k stars 102 forks source link

Just using the segmentation part #109

Open ayushnangia opened 2 months ago

ayushnangia commented 2 months ago

Hi I was looking to just fine tune the segmentation part for my down stream task. Could you please let me know how to get that as both object detection and segmentation are there

And is there a way to just get the bit mask of the segmentation or do i need to change the code for that in some files?

lastly, what would be the best config for fine tuning models for mask dino?

FengLi-ust commented 2 months ago

Hi, I think you only need to finetune the decoder part, it is lightweight and contains both detection and segmentaiton.