Visual Attention Model Based Visual Hierarchy Recommendation Framework for Improved User Experience Design Dataset

A dataset comprising videos recorded through the Gaze Cloud Eye Tracker, a sophisti- cated eye tracking system is used. Each video in the dataset includes two crucial visual components: a heatmap and a circle pointer, illustrating the user’s eye movement dur- ing the recording. Calibration is done before each video session to address any potential errors. The calibration process involves a thorough review of the user’s setup including the monitor dimensions, the distance at which the user is from the screen and the po- sition of his eyeballs and gaze. This calibration process emphasizes the precision and dependability of the captured eye movement data. Each video within the dataset cap- tures the user’s browsing activity on a website. This unintentional browsing scenario is intended to simulate a more natural and free-flowing user experience, allowing for a comprehensive analysis of eye movements in a real-world web browsing context.

本数据集包含通过Gaze Cloud眼动仪（Gaze Cloud Eye Tracker）采集的视频，该设备为一款先进的眼动追踪系统。数据集中的每段视频均包含两个核心视觉组件：热图与圆形指针，用于直观展示录制过程中用户的眼动轨迹。每段视频录制前均会完成校准流程，以规避潜在误差。校准环节会全面核查用户的实验设置，包括显示器尺寸、用户与屏幕的距离，以及用户眼球与视线的位置。该校准流程旨在保障采集到的眼动数据具备高精度与可靠性。数据集中的每段视频均记录了用户在某一网站的浏览行为，该场景为非刻意浏览模式，旨在模拟更自然、更流畅的用户体验，从而能够在真实的网页浏览场景中实现对眼动行为的全面分析。