2 verified, source-linked AI implementations documented between 2022–2023.
Readers click "kuula artiklit" to hear articles read aloud using Neurokõne neural speech synthesis model (TartuNLP, University of Tartu); voice """"Meelis""""";" 10 speaker voices available""
Real-time AI-generated subtitles for all live ETV and ETV2 broadcasts including news; Kiirkirjutaja (""""fast writer"""") system developed by TalTech"; Kaldi-based TDNN-F speech models;" available as accessibility/subtitle channel for hard-of-hearing viewers""
ERR is one of 4 organizations in Estonia documented in the index, which holds 9 cases from the country. Its work spans 1 function type, including Accessibility, documented across 9 organizations worldwide.