Binaural Rendering Dataset

该数据集包含了在受控环境中录制的配对式的立体声和双耳信号，旨在促进双耳渲染研究。使用H3-VR进行立体声信号的录制，同时采用Neumann KU100假人头进行双耳信号的捕捉。数据被分割成不重叠的一分钟片段，以便于训练和评估。规模上，训练集包含31分钟的数据，评估集包含18分钟的数据。该任务的目标是对立体声信号进行双耳渲染。

This dataset contains paired stereo and binaural signals recorded in a controlled environment, aiming to advance binaural rendering research. Stereo signals were recorded using an H3-VR, while binaural signals were captured with a Neumann KU100 dummy head. The data is split into non-overlapping one-minute segments to facilitate training and evaluation. The training set includes 31 minutes of data, and the evaluation set contains 18 minutes of data. The goal of this task is to perform binaural rendering on stereo signals.