@inproceedings{silva-etal-2026-software,
    title = "Software for Automatic Speech Recognition via Whisper models applied to Oral History interviews in the {P}ortuguese language",
    author = "Silva, Edgleide de Oliveira Clemente da  and
      Zagatti, Fernando Rezende  and
      Lopes, Filipe Loyola  and
      Duarte, Anderson Dias  and
      Bonacin, Rodrigo  and
      Alves, Angela Maria",
    editor = "Souza, Marlo  and
      de-Dios-Flores, Iria  and
      Santos, Diana  and
      Freitas, Larissa  and
      Souza, Jackson Wilke da Cruz  and
      Ribeiro, Eug{\'e}nio",
    booktitle = "Proceedings of the 17th International Conference on Computational Processing of {P}ortuguese ({PROPOR} 2026) - Vol. 1",
    month = apr,
    year = "2026",
    address = "Salvador, Brazil",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-dnd/2026.propor-1.30/",
    pages = "302--310",
    ISBN = "979-8-89176-387-6",
    abstract = "This paper presents Ethos AT, a desktop software for automatic transcription that uses OpenAI Whisper models, enabling local processing and ensuring data privacy and accessibility for users who are not necessarily programming experts, such as oral history researchers. A comparative analysis of six Whisper models (small, medium, large, large-v2, large-v3, and turbo) was conducted to analyze performance in terms of transcription accuracy, error types, and processing time. Results indicate that larger models achieve higher lexical accuracy, while smaller ones provide faster execution with acceptable quality for general use; the turbo model showed an effective balance between accuracy and speed. Overall, Ethos AT offers a secure, efficient, and user-friendly solution for academic and institutional contexts."
}