the split of dataset
Closed this issue · 2 comments
Zizzzzzzz commented
In paper, the split of dataset is 30 sequences for training, 3 sequences validation, and 8 sequences testing. But in DSEC_detection_labels folder, the split of dataset is 33 sequences for training, 3 sequences validation, and 5 sequences testing. I am confused.
abhishek1411 commented
The 8 sequences for testing in the paper were divided into 5 sequences for testing and 3 sequences for validation when the labels were released publicly. You could use the entire 8 sequences for testing and split a part of the training set into validation for hyperparameter tuning.
Zizzzzzzz commented
OK. Thank you!