A Dataset for Automatic Assessment of TTS Quality in Spanish

HablaMOS: A Dataset for Automatic Assessment of TTS Quality in SpanishLink to Paper! Data CollectionTTS systems include neural, concatenative, parametric, and proprietary models covering various dialects. Real human samples serve as high-quality references.Two data augmentation methods (Vocal Tract Length Perturbation and Griffin-Lim phase alteration) were applied to increase diversity. They are identified by "_VTLP" and "_GL" suffixes. If you use this dataset, please cite the following paper: A Dataset for Automatic Assessment of TTS Quality in Spanish Alejandro Sosa Welford, Leonardo Pepino https://arxiv.org/abs/2507.01805