Closed flovst closed 1 year ago
The implementation in pytorch/model/detector/regression_loss (line 144) seems to apply x (instead of x_abs) in the case selection. Simply changing is_negative_branch = (x < (1.0 / sigma_squared)) to is_negative_branch = (x_abs ...). shall fix it.
pytorch/model/detector/regression_loss (line 144)
is_negative_branch = (x < (1.0 / sigma_squared))
is_negative_branch = (x_abs ...).
Thank you for this project!
Good eye, thank you! Fixed.
trzy
commented
1 year ago
The implementation in
pytorch/model/detector/regression_loss (line 144)seems to apply x (instead of x_abs) in the case selection. Simply changingis_negative_branch = (x < (1.0 / sigma_squared))tois_negative_branch = (x_abs ...).shall fix it.Thank you for this project!