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
Hello,
Thank you for open-sourcing this work. I have a few questions regarding it. When I tried to reproduce this work on a similar snowykitti dataset, I found that I was able to train the model normally, but when using the trained weights for evaluation, all predictions were incorrect, with both TP and FN being 0. Interestingly, when I used a model pre-trained with WADS without changing anything else, I was able to get reasonable results, although the recall also tended towards 0.
The text was updated successfully, but these errors were encountered:
Unfortunately, this repo has not been tested on SnowKITTI.
Interestingly, when I used a model pre-trained with WADS without changing anything else, I was able to get reasonable results, although the recall also tended towards 0.
Can you post some example visualizations showing positive & negative points separately?
Hello,
Thank you for open-sourcing this work. I have a few questions regarding it. When I tried to reproduce this work on a similar snowykitti dataset, I found that I was able to train the model normally, but when using the trained weights for evaluation, all predictions were incorrect, with both TP and FN being 0. Interestingly, when I used a model pre-trained with WADS without changing anything else, I was able to get reasonable results, although the recall also tended towards 0.
The text was updated successfully, but these errors were encountered: