Learning Expressive Prompting With Residuals for Vision Transformers
This work addresses efficient adaptation of vision transformers for tasks like image classification, few-shot learning, and semantic segmentation, representing an incremental improvement in prompt learning methods.
The paper tackled the problem of adapting vision transformers efficiently by introducing Expressive Prompts with Residuals (EXPRES), which achieved state-of-the-art prompt tuning on all three categories of the VTAB benchmark and was an order of magnitude more prompt efficient than existing baselines.
Prompt learning is an efficient approach to adapt transformers by inserting learnable set of parameters into the input and intermediate representations of a pre-trained model. In this work, we present Expressive Prompts with Residuals (EXPRES) which modifies the prompt learning paradigm specifically for effective adaptation of vision transformers (ViT). Out method constructs downstream representations via learnable ``output'' tokens, that are akin to the learned class tokens of the ViT. Further for better steering of the downstream representation processed by the frozen transformer, we introduce residual learnable tokens that are added to the output of various computations. We apply EXPRES for image classification, few shot learning, and semantic segmentation, and show our method is capable of achieving state of the art prompt tuning on 3/3 categories of the VTAB benchmark. In addition to strong performance, we observe that our approach is an order of magnitude more prompt efficient than existing visual prompting baselines. We analytically show the computational benefits of our approach over weight space adaptation techniques like finetuning. Lastly we systematically corroborate the architectural design of our method via a series of ablation experiments.