anthonyweidai / SvANet

Official implementation of "Exploiting Scale-Variant Attention for Segmenting Small Medical Objects"
MIT License
37 stars 2 forks source link

Using tinyObjectsValidation.py for reasoning will result in an error,Can anyone give me some guidance? #6

Closed GOfBlue closed 1 month ago

GOfBlue commented 2 months ago

Traceback (most recent call last): File "/mnt/data12t/gyc_data_new/SvANet-main/tools/tinyObjectsValidation.py", line 12, in from lib.data import getMyDataset File "/mnt/data12t/gyc_data_new/SvANet-main/lib/data/init.py", line 2, in from .mydatasetReg import getMyDataset File "/mnt/data12t/gyc_data_new/SvANet-main/lib/data/mydatasetReg.py", line 36, in importModule(file, SkipFolder=['collate_fn', 'process', 'sampler', 'tools']) File "/mnt/data12t/gyc_data_new/SvANet-main/lib/utils/utils.py", line 14, in importModule SubFolds.extend(next(os.walk(RegisterPath))[1]) # return folder only list StopIteration ERROR conda.cli.main_run:execute(49): conda run python /mnt/data12t/gyc_data_new/SvANet-main/tools/tinyObjectsValidation.py failed. (See above for error)

anthonyweidai commented 2 months ago

You should move the tinyObjectsValidation.py to the working directory.