We propose a method, Continual Prompted Transformer for Test-Time Training (CPT4), that enhances the Vision Transformer (ViT) model by incorporating shared prompts (small learnable parameters) and a batch normalization module, aiming to mitigate catastrophic forgetting and handle domain shifts ...
A beautiful and useful prompt for your shell Python 6,182 745 Updated Mar 19, 2024 lukas-blecher / LaTeX-OCR pix2tex: Using a ViT to convert images of equations into LaTeX code. Python 11,157 926 Updated Jun 5, 2024 akaalias / obsidian-extract-pdf-highlights Extract highlights...