ExPLoRA: Parameter-Efficient Extended Pre-Training to Adapt Vision Transformers under Domain Shifts.

Samar Khanna, Medhanie Irgau, David B. Lobell, Stefano Ermon

Browse the full ICML paper archive.