WiDESPREaD

WiDESPREaD数据集由法国雷恩中央理工学院创建，包含1005对耳形和相关的pinna-related transfer functions（PRTFs）。该数据集通过随机耳形绘制和声学模拟生成，用于增强和降低头相关传递函数（HRTF）的维度。数据集的创建旨在解决HRTF个性化问题，特别是在双耳合成中的应用，通过提供多样化的耳形和PRTFs，帮助研究者更好地理解和模拟个体间的差异。

The WiDESPREaD dataset was developed by École Centrale de Rennes, France, which contains 1005 pairs of ear shapes and their corresponding pinna-related transfer functions (PRTFs). This dataset is generated through random ear shape generation and acoustic simulation, and is utilized for dimensionality enhancement and reduction of head-related transfer functions (HRTFs). The dataset was created to address the HRTF personalization problem, particularly for applications in binaural synthesis. By providing diverse ear shapes and PRTFs, it helps researchers better understand and simulate inter-individual differences.