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
Now ,I Find it so so so difficult to do more things in just few days.
Maybe this competition want us to do the sim2real work owing that the official don't support real ojbects,and also the pose data is so difficult to get(LabelFusion can solve with real objects).So I think it may work through sim2real in the future(but it may not work now for we has little time left)
Others ways I think it may work is just simply use traditional way to classify objects ,get grasp rect to pick(Shown in picture) and write more if else to solve the problem.But I don't think it is meaningful except for the competition.
So if it is possible ,I hope we can communicate more.
My e-mail is :1206413225@qq.com(the same as qq),if it is possible ,I want to communicate with anyone to get new ideas.I really like grasping and want to make friends with everyone.
Best Wishes.
The text was updated successfully, but these errors were encountered:
We did find all the real objects on TaoBao but being in Australia are not able to source in time.
We had a similar thought with traditional methods while interesting they aren't really applicable outside of this competition.
Thanks for the input. We welcome any kind of inspiration, this will help us better conduct the competition in the future. If you have specific suggestions, please write to us at info@ocrtoc.org.
We did find all the real objects on TaoBao but being in Australia are not able to source in time.
We had a similar thought with traditional methods while interesting they aren't really applicable outside of this competition.
Yeah,Except for the competition,traditional methods I think they have no meaning.But this is what I can handle method...
Now ,I Find it so so so difficult to do more things in just few days.
Maybe this competition want us to do the sim2real work owing that the official don't support real ojbects,and also the pose data is so difficult to get(LabelFusion can solve with real objects).So I think it may work through sim2real in the future(but it may not work now for we has little time left)
Others ways I think it may work is just simply use traditional way to classify objects ,get grasp rect to pick(Shown in picture) and write more if else to solve the problem.But I don't think it is meaningful except for the competition.
So if it is possible ,I hope we can communicate more.
My e-mail is :1206413225@qq.com(the same as qq),if it is possible ,I want to communicate with anyone to get new ideas.I really like grasping and want to make friends with everyone.
Best Wishes.
The text was updated successfully, but these errors were encountered: