Hello respected author,
I noticed that in many parts of the code, components from the fsdet library are imported, such as in train_net.py:
from fsdet.config import get_cfg, set_global_cfg
from fsdet.engine import DefaultTrainer, default_argument_parser, default_setup
from fsdet.evaluation import (
COCOEvaluator, DatasetEvaluators, LVISEvaluator, PascalVOCDetectionEvaluator, verify_results
)
However, equivalent definitions for these components also exist within the iFS-RCNN codebase. Should I replace all imports from fsdet with the corresponding implementations from iFS-RCNN? I am confused about why the iFS-RCNN code defines these modules but does not use them directly. Could you kindly explain the rationale behind this design choice? Thank you very much for your time and guidance.
Hello respected author,
I noticed that in many parts of the code, components from the fsdet library are imported, such as in train_net.py:
from fsdet.config import get_cfg, set_global_cfg
from fsdet.engine import DefaultTrainer, default_argument_parser, default_setup
from fsdet.evaluation import (
COCOEvaluator, DatasetEvaluators, LVISEvaluator, PascalVOCDetectionEvaluator, verify_results
)
However, equivalent definitions for these components also exist within the iFS-RCNN codebase. Should I replace all imports from fsdet with the corresponding implementations from iFS-RCNN? I am confused about why the iFS-RCNN code defines these modules but does not use them directly. Could you kindly explain the rationale behind this design choice? Thank you very much for your time and guidance.