One Block, Multiple Depths: Recurrent Vision Transformers with Depth-Programmed Experts
By Adrian Bulat · Paper · cs.CV
In this work, we show that a single Transformer block, applied recurrently, can match the accuracy of a full-depth vision encoder at comparable inference FLOPs without intermediate feature distillation. reViT restores depth-specific transformations by representing the FFN at each
Tags: Cs.cv