Hi
I was wondering that for training our custom dataset, should test images/labels names be same as train images/labels names? Because I one of my train label has the name 0000.txt, but test label has no such name. However, I get the following error:
Traceback (most recent call last):
File "test.py", line 139, in <module>
args.debugDec,)
File "test.py", line 79, in test
debugEnc=debugEnc, debugDec=debugDec)
File "/home/ubuntu/lstr-custom/LSTR/test/custom.py", line 225, in testing
repeat=repeat, isEncAttn=debugEnc, isDecAttn=debugDec)
File "/home/ubuntu/lstr-custom/LSTR/test/custom.py", line 216, in kp_detection
eval_result = evaluator.eval(label='{}'.format(os.path.basename(exp_name)))
File "/home/ubuntu/lstr-custom/LSTR/db/utils/evaluator.py", line 28, in eval
return self.dataset.eval(self.exp_dir, self.predictions, self.runtimes, **kwargs)
File "/home/ubuntu/lstr-custom/LSTR/db/custom.py", line 352, in eval
return f1_metric.eval_predictions(self.anno_root, eval_dir, width=30, official=True, sequential=False)
File "/home/ubuntu/lstr-custom/LSTR/db/utils/f1_metrics.py", line 116, in eval_predictions
annotations = load_fld_gt(anno_dir, names)
File "/home/ubuntu/lstr-custom/LSTR/db/utils/f1_metrics.py", line 107, in load_fld_gt
anno_data = load_culane_img_data(anno_file)
File "/home/ubuntu/lstr-custom/LSTR/db/utils/f1_metrics.py", line 84, in load_culane_img_data
with open(path, 'r') as data_file:
**FileNotFoundError: [Errno 2] No such file or directory: './raws/test_labels/0000.txt**'
Hi I was wondering that for training our custom dataset, should test images/labels names be same as train images/labels names? Because I one of my train label has the name 0000.txt, but test label has no such name. However, I get the following error: