pMoE: Prompting Diverse Experts Together Wins More in Visual Adaptation

TL;DR

The paper introduces pMoE, a novel prompt tuning method that combines multiple domain experts through dynamic dispatching, significantly improving visual adaptation performance across diverse tasks while maintaining computational efficiency.

Contribution

pMoE is the first to integrate multiple domain-specific experts via prompt tokens and dynamic dispatching, enhancing versatility and performance in visual adaptation tasks.

Findings

Achieves superior performance across 47 tasks.

Outperforms existing methods with significant margins.

Balances computational efficiency with adaptation effectiveness.

Abstract

Parameter-efficient fine-tuning has demonstrated promising results across various visual adaptation tasks, such as classification and segmentation. Typically, prompt tuning techniques have harnessed knowledge from a single pre-trained model, whether from a general or a specialized medical domain. However, this approach typically overlooks the potential synergies that could arise from integrating diverse domain knowledge within the same tuning process. In this work, we propose a novel Mixture-of-Experts prompt tuning method called pMoE, which leverages the strengths of multiple expert domains through expert-specialized prompt tokens and the learnable dispatcher, effectively combining their expertise in a unified model framework. Our pMoE introduces expert-specific prompt tokens and utilizes a dynamic token dispatching mechanism at various prompt layers to optimize the contribution of each domain expert during the adaptation phase. By incorporating both domain knowledge from diverse experts, the proposed pMoE significantly enhances the model's versatility and applicability to a broad spectrum of tasks. We conduct extensive experiments across 47 adaptation tasks, including both classification and segmentation in general and medical domains. The results demonstrate that our pMoE not only achieves superior performance with a large margin of improvements but also offers an optimal trade-off between computational efficiency and adaptation effectiveness compared to existing methods.