Tool-IQA: Augmenting Image Quality Assessment with Simple Tools
Researchers have developed a new method called Tool-IQA that enhances image quality assessment by equipping Vision-Language Models (VLMs) with interactive tools. Unlike static one-shot scoring, Tool-IQA utilizes a Magnifier for detailed local inspection and a Gamma Corrector to reveal hidden artifacts. This augmented workflow, which includes initial observation, tool-assisted inspection, and final scoring, significantly outperforms existing state-of-the-art models on benchmarks like the CLIVE dataset. AI
IMPACT Introduces a novel approach to image quality assessment by integrating interactive tools with VLMs, potentially improving objective image evaluation.