A11y-Compressor: A Framework for Enhancing the Efficiency of GUI Agent Observations through Visual Context Reconstruction and Redundancy Reduction

TL;DR

A11y-Compressor is a framework that transforms accessibility trees into compact, structured representations, reducing input size and improving AI agent performance in GUI tasks.

Contribution

It introduces a lightweight transformation pipeline that reduces redundancy and adds structure to accessibility trees for better GUI understanding by AI agents.

Findings

Reduces input tokens to 22% of original size.

Improves task success rates by 5.1 percentage points on average.

Enhances GUI observation efficiency for AI agents.

Abstract

AI agents that interact with graphical user interfaces (GUIs) require effective observation representations for reliable grounding. The accessibility tree is a commonly used text-based format that encodes UI element attributes, but it suffers from redundancy and lacks structural information such as spatial relationships among elements. We propose A11y-Compressor, a framework that transforms linearized accessibility trees into compact and structured representations. Our implementation, Compressed-a11y, applies a lightweight and structured transformation pipeline with modal detection, redundancy reduction, and semantic structuring. Experiments on the OSWorld benchmark show that Compressed-a11y reduces input tokens to 22% of the original while improving task success rates by 5.1 percentage points on average.