Interface icons are prevalent on various digital applications. However, designers often rely on informal evaluation to determine the usability of icons due to limited time and budgets, thus often resulting in low usability icons. In this paper, we propose a unique human-in-the-loop framework that allows users to improve the usabilities of interface icons efficiently. We formulate several usability criteria into a perceptual usability function and enable users to iteratively revise an icon set with an interactive design tool, EvIcon. During the revision process, our design tool provides two types of instant perceptual usability feedback. First, we provide perceptual usability feedback modeled by deep learning models trained on collected crowdsourced perceptual ratings. Second, we learned an embedding space of a large icon dataset to assist users in improving icons’ visual distinguishability among icons within the user-prepared icon set. To provide the perceptual prediction, we compiled IconCEPT10K, the first large-scale dataset of perceptual usability ratings over 10000 interface icons by conducting a crowdsourcing study. We demonstrated that our framework can benefit the interface icon revision process for UI designers with a wide range of professional experience. Moreover, the interface icons designed using our framework achieved better semantic distance and familiarity verified by an additional online user study.

@article{ shen2023evicon,	      
author    = {I-Chao Shen and Fuyin Cherng and Yu-Ting Wang and Takeo Igarashi and Wen-Chieh Lin and Bing-Yu Chen},
title     = {EvIcon: Designing High-Usability Icon with Human-in-the-loop Exploration},
journal   = {Computer Graphics Forum},
year      = {2023}
}