Toward learning human-aligned cross-domain robust models by countering misaligned features.

Haohan Wang, Zeyi Huang, Hanlin Zhang, Yong Jae Lee, Eric P. Xing

Browse the full UAI paper archive.