COLORA: Efficient Fine-Tuning for Convolutional Models with a Study Case on Optical Coherence Tomography Image Classification
This provides a parameter-efficient fine-tuning method for medical image classification, but it is incremental as it extends LoRA to CNNs.
They tackled efficient fine-tuning for convolutional neural networks by introducing CoLoRA, which reduces trainable parameters to 0.2 compared to conventional methods and achieves up to 1% accuracy and 0.013 AUC improvements on OCTMNISTv2 while cutting per-epoch training time by nearly 20%.
We introduce CoLoRA (Convolutional Low-Rank Adaptation), a parameter-efficient fine-tuning method for convolutional neural networks (CNNs). CoLoRA extends LoRA to convolutional layers by decomposing kernel updates into lightweight depthwise and pointwise components.This design reduces the number of trainable parameters to 0.2 compared to conventional fine-tuning, preserves the original model size, and allows merging updates into the pretrained weights after each epoch, keeping inference complexity unchanged. On OCTMNISTv2, CoLoRA applied to VGG16 and ResNet50 achieves up to 1 percent accuracy and 0.013 AUC improvements over strong baselines (Vision Transformers, state-space, and Kolmogorov Arnold models) while reducing per-epoch training time by nearly 20 percent. Results indicate that CoLoRA provides a stable and effective alternative to full fine-tuning for medical image classification.