UICrit
收藏资源简介:
UICrit数据集由加州大学伯克利分校的研究团队创建,包含3,059条自然语言设计批评和质量评级,针对983个移动UI屏幕。数据集的创建过程涉及七位经验丰富的设计师,他们根据既定的设计指南和个人的设计经验提供批评和评级。每个批评包括对UI屏幕相关区域的边界框标注,以及美学、可用性和整体设计质量的数值评级。数据集的应用领域主要是在自动化设计评估中,通过提供高质量的设计反馈来改进LLM生成的UI反馈,从而提高设计效率和质量。
The UICrit dataset was developed by a research team at the University of California, Berkeley, and consists of 3,059 natural language design critiques and quality ratings for 983 mobile UI screens. The dataset creation process involved seven experienced designers, who provided critiques and ratings based on established design guidelines and their individual design expertise. Each critique includes bounding box annotations for the relevant regions of the corresponding UI screen, alongside numerical ratings for aesthetics, usability, and overall design quality. The primary application domain of this dataset is automated design evaluation, where it can provide high-quality design feedback to improve the UI feedback generated by Large Language Models (LLMs), thereby enhancing design efficiency and quality.

- 1UICrit: Enhancing Automated Design Evaluation with a UICritique Dataset加州大学伯克利分校 · 2024年



