search

cviviers / YOLOv5-6D-Pose

6-DoF Pose estimation based on the YOLOv5 framework. Specific focus on instruments in X-ray applications
https://ieeexplore.ieee.org/document/10478293
GNU General Public License v3.0
32 stars 6 forks source link

image miss lables and no masks found #5

Closed hippohi closed 1 month ago

hippohi commented 2 months ago

hi dear author ,i tried to run test.py then i got this: Fusing layers... Model Summary: 391 layers, 97574820 parameters, 0 gradients, 225.4 GFLOPS val: Scanning 'D:\data\LINEMOD\cat\labels' for images, labels and masks... 1179 found, 0 missing, 0 empty, 0 corrupted, 0 found, 1179 missing, 0 empty, 0 corrupted: 100%|██████████| 1179/1179 [00:01<00:00, 1101.94it/s] val: WARNING: No masks found in D:\data\LINEMOD\cat\labels.cache.

i dont know why the lables are all missing and are those masks scanned automatically just like the lables? besides "--rect" and "--cache" aren`t existing in parser.add_argument so i deleted them.

cviviers commented 2 months ago

Hey @hippohi,

Could it be that in a previous run, when the cache was created, there was something else wrong? Do you mind deleting the cache file and running it again? There will be explicit error messages if anything is indeed wrong.

hippohi commented 1 month ago

Hey @hippohi,

Could it be that in a previous run, when the cache was created, there was something else wrong? Do you mind deleting the cache file and running it again? There will be explicit error messages if anything is indeed wrong.

hey my friend helped me solve the issue ! If you run the train.py on WINDOWS,you may need to change the "replace('/00', '/')" into "replace('\\00', ' \\')" which is located under the function "def Linemodimg2mask_paths(img_paths):" of dataset.py.