Detectron: pre-trained weights from coco dataset
hello, but do you also have pretrained weights from coco dataset instead of imagenet?
thanks
coldgemini
All 3 comments
Proposal, Box, and Mask Detection Baselines
These models are all trained on coco.
ppwwyyxx
on 27 Jan 2018
yes, but my dataset has class number different from coco dataset so your mentioned model cannot be directly loaded into the training scripts. I don't know how caffe2 handles model importing with different layer sizes. I know however in caffe I can just modify the layer name.
coldgemini
on 27 Jan 2018
To further fine-tune a COCO model on another dataset (e.g., Cityscapes as we did), you need to remove the layers that are class-dependent (in Mask R-CNN, in bbox cls, bbox reg, and mask pred layers), and reinitialize them.
In Detectron, the weights are loaded by matching names, so a simple way to do this is either to change the name of the class-dependent layers when you do further fine-tuning, or add a flag in the weight loading to not load these weights.
KaimingHe
on 27 Jan 2018
Related issues
elfpattern
·
3Comments
realwecan
·
3Comments
junxiaoge
·
3Comments
Emma0928
·
3Comments
rbgirshick
·
3Comments
Most helpful comment
To further fine-tune a COCO model on another dataset (e.g., Cityscapes as we did), you need to remove the layers that are class-dependent (in Mask R-CNN, in bbox cls, bbox reg, and mask pred layers), and reinitialize them.
In Detectron, the weights are loaded by matching names, so a simple way to do this is either to change the name of the class-dependent layers when you do further fine-tuning, or add a flag in the weight loading to not load these weights.