OWOD
OWOD copied to clipboard
JosephKJ
(CVPR 2021 Oral) Open World Object Detection
Can the code for splitting the data be provided? I am aiming to split our custom dataset. But there is still some confusion after reading the paper. To be specific,...
Thanks for nice paper and releasing source code. Unfortunately, i got confused with your description about install. I see that you modified original detectron2 package. (i.e. dataset define) Thus, there...
Why not dense unknown boxes? In the "A sample qualitative result" example, apples and oranges are marked as unknown to begin with, but why aren't glasses, human ears, eyebrows, hair,...
Hi Joseph, I have a few questions in regards to the data split. Could you please solve my confusion? 1. I can see that all data split codes are in...
Thank you for your reply @JosephKJ Here are two questions I still in trouble： When I finish task2 training, is the model_final.pth in t2_final works better than that in t2_ft?...
Thank you for your nice paper. But I don't know how to conduct incremental object detection. - According to the open world setting. When the model labels sufficient unknown classes,...
Before you draw the 'energy.png', you analysis the unknown and known by using Fit_Weibull_3P. wb_unk = Fit_Weibull_3P(failures=unk, show_probability_plot=False, print_results=False) wb_known = Fit_Weibull_3P(failures=known, show_probability_plot=False, print_results=False) I want to know why you...
May I ask where is the code of Top-k background marked as "unknown" ? I can not find it
In each test task and val task, how do you determine the number of unknown targets? Is this target not pre-labeled? Do you also use the unknown annotated RPN during...
Hi guys！ Have you encountered this problem before? How did you solve it? My environment is Python 3.8+torch1.10+cuda113；reliability == 0.5.6 Thank you!