Finetuning DINOv2 with LoRA for Image Segmentation
Finetuned DINOv2 encoder weights using Low-Rank Adaptation (LoRA) and a 1×1 convolution decoder, achieving 74.6% Mean IoU on Pascal VOC with only 0.69% trainable parameters.
Finetuned DINOv2 encoder weights using Low-Rank Adaptation (LoRA) and a 1×1 convolution decoder, achieving 74.6% Mean IoU on Pascal VOC with only 0.69% trainable parameters.