# Large-scale image feature extraction

AIMv2 is an image encoder model based on the timm library, adopting the extra-large patch14_448 architecture, suitable for image feature extraction tasks.

Segment-Anything Vision Transformer (SAM ViT) image feature model, which only includes feature extraction and fine-tuning capabilities, without the segmentation head.

RegNetY-128GF feature extraction model, pre-trained using the SEER method on 2 billion random web images with self-supervised learning

RegNetY-32GF feature extraction model, pretrained on 2 billion random web images using SEER method, suitable for image classification and feature extraction tasks