You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, why did you choose to set lambda to 1 when calculating the total loss?
I know that in the FOTS paper they set lambda to 1, but in other FOTS repos the value is often 0,01.
Do you know in which interval the values of the recognition loss and the detection loss are?
The text was updated successfully, but these errors were encountered:
Hi, @daniiki. That's a good question!
I was wondering why everyone used 0.01 for lambda instead of 1, which is what the paper says.
So I changed it to 1 to see how the model would perform, but I didn't get better result. In fact, I got much worse detection results as you can guess.
So if you want to get good detection results, you definitely should set lambda = 0.01 as everyone sets.
I was thinking I should reset the value to the original value, but I totally forgot to update it and push it to the repo.
OK. I will set it back now then.
Hi, why did you choose to set lambda to 1 when calculating the total loss?
I know that in the FOTS paper they set lambda to 1, but in other FOTS repos the value is often 0,01.
Do you know in which interval the values of the recognition loss and the detection loss are?
The text was updated successfully, but these errors were encountered: