ebellob/voxpopuli_spanish_enhanced

该数据集是facebook/voxpopuli数据集中西班牙语子集的经过处理和增强的版本，专为支持高质量的语音研究而设计，包括文本到语音（TTS）、自动语音识别（ASR）以及语音增强和鲁棒性研究。数据集通过CleanUNet进行语音去噪，并通过FlashSR进行音频超分辨率处理，以提高音频质量。数据集包含50,922个样本，大小约为105 GB，所有音频均为西班牙卡斯蒂利亚（半岛西班牙语）方言，源自欧洲议会的全体会议录音。

This dataset is a processed and enhanced version of the Spanish subset of the facebook/voxpopuli dataset, designed to support high-quality speech research, including Text-to-Speech (TTS), Automatic Speech Recognition (ASR), and speech enhancement and robustness studies. The dataset has been denoised using CleanUNet and super-resolved using FlashSR to improve audio quality. It contains 50,922 samples with an approximate size of 105 GB, all in Castilian (Peninsular) Spanish dialect sourced from European Parliament plenary sessions.