AIDE: Agentically Improve Visual Language Model with Domain Experts

TL;DR

AIDE is a framework that allows visual language models to autonomously improve by leveraging domain experts, enabling performance gains without larger models or human supervision.

Contribution

We introduce AIDE, a novel autonomous framework that enhances VLMs using domain expert models, bypassing the need for larger models or human involvement.

Findings

Achieves performance improvements on multiple benchmarks.

Operates without reliance on larger models or human supervision.

Provides a scalable, resource-efficient improvement method.

Abstract

The enhancement of Visual Language Models (VLMs) has traditionally relied on knowledge distillation from larger, more capable models. This dependence creates a fundamental bottleneck for improving state-of-the-art systems, particularly when no superior models exist. We introduce AIDE (Agentic Improvement through Domain Experts), a novel framework that enables VLMs to autonomously enhance their capabilities by leveraging specialized domain expert models. AIDE operates through a four-stage process: (1) identifying instances for refinement, (2) engaging domain experts for targeted analysis, (3) synthesizing expert outputs with existing data, and (4) integrating enhanced instances into the training pipeline. Experiments on multiple benchmarks, including MMMU, MME, MMBench, etc., demonstrate AIDE's ability to achieve notable performance gains without relying on larger VLMs nor human supervision. Our framework provides a scalable, resource-efficient approach to continuous VLM improvement, addressing critical limitations in current methodologies, particularly valuable when larger models are unavailable to access.