Unseen_Object_Pose icon indicating copy to clipboard operation
Unseen_Object_Pose copied to clipboard

Published 20 hours ago • verified publisher icon sailor-z

About The TLESS result

Open qifengluo opened this issue 1 year ago • 1 comments

Hi, I have reviewed the test results of the paper and found that there are significant differences between the MPE data in the evaluation table of the less dataset and the original text, and the evaluation methods are also different (train joint obj1-18, test 19-30). Is the MPE dataset in the article the result of your own testing? And MPE, DISP, and EEGP-AAE use the same evaluation strategy, and the same method is not used for comparison in the article

qifengluo avatar Jan 12 '24 04:01 qifengluo

Sorry, I forgot the paper name. DISP: Disentangled Implicit Shape and Pose Learning for Scalable 6D Pose Estimation MPE：Multi-path Learning for Object Pose Estimation Across Domains. EEGP-AAE is a method for seen objects, do not related to above.

qifengluo avatar Jan 12 '24 08:01 qifengluo