Mittag, Gabriel2021-04-192021-04-192021-04https://depositonce.tu-berlin.de/handle/11303/13012.5http://dx.doi.org/10.14279/depositonce-11807.5The NISQA Speech Quality Corpus contains two training, two validation and four test datasets that have been used for training and evaluating the deep learning based speech quality model NISQA. Overall, it contains more than 14,000 speech files with annotated speech quality and speech quality dimension ratings.en620 Ingenieurwissenschaften und zugeordnete Tätigkeitenspeech qualitydeep learningspeech samplesmachine learningquality of experienceAudio