LoCA: Spatially-Aware Low-Rank Convolutional Adaptation of Vision Foundation Models
Low-Rank Convolutional Adaptation (LoCA), a convolution-aware PEFT framework that addresses spatial-channel entanglement by decoupling channel and spatial adaptation and achieves competitive or state-of-the-art performance across fine-grained classification, domain-generalized semantic segmentation, and generative benc...