3D ViT-L encoders (SimMIM on LMD, from DINOv3 LVD-1689M or from scratch) and their affinity fine-tunes for neuron segmentation in EM and ExM at 8 nm.
-
ai-hhmi/lmd-vitl-p8-simmim-lvd
Image Feature Extraction • 0.3B • Updated • 3.33k -
ai-hhmi/lmd-vitl-p8-simmim-scratch
Image Feature Extraction • 0.3B • Updated • 3.32k -
ai-hhmi/lmd-vitl-p8-affinity-lvd
Image Segmentation • 0.3B • Updated • 90 -
ai-hhmi/lmd-vitl-p8-affinity-scratch
Image Segmentation • 0.3B • Updated • 69