Open xxxxxxxx-dl opened 7 years ago
In the source code of tools/rfcn_alt_opt_5stage_ohem.py, stage 2 rfcn training uses initial net weight from initial imagenet model. Why didn't it use net weight trained from stage 2 rpn but init from imagenet model?
In the source code of tools/rfcn_alt_opt_5stage_ohem.py, stage 2 rfcn training uses initial net weight from initial imagenet model. Why didn't it use net weight trained from stage 2 rpn but init from imagenet model?