A carregar...
Publicação
Nonperiodic pathologic voice signals classification using mel-spectrogram and VGGish
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 251.64 KB | Adobe PDF |
Orientador(es)
Resumo(s)
In this work and the literature, voice signals can be classified as peri-odic (type 1) or either some periodicity (type 2) and chaos (type 3). This work aims to classify signs into types 1, 2 or 3 to be subsequently applied in a classifi-cation system for pathological/control signs. The original dataset is composed of 466 type 1 individuals, 900 type 2 individuals, and 84 type 3 individuals classi-fied by an otolaryngologist. 15% of the data was used for testing and the remain-ing 85% was used for training and validation. A data augmentation technique was applied to balance the data in training set. Therefore, for the test set, 3380 sounds were used, 1020 type 1, 1280 type 2 and 1080 type 3. Of these, 80% were used for training and 20% for validation. The Mel spectrograms of the signals were used in the input of a VGGish to retrain the model in classifying the 3 types of signals. Regarding test accuracy, this network obtained 71.2%.
Descrição
Palavras-chave
Contexto Educativo
Citação
Fernandes, Joana; Pinto, João; Moura, Carla; Vilarinho, Helena; Teixeira, Felipe; Freitas, D.; Teixeira, João Paulo (2025) Nonperiodic pathologic voice signals classification using mel-spectrogram and VGGish. In International Conference on Demographic Transition, Health and Technologies, ICDTHT 2025. p. 3-13. ISBN 978-303194900-5. DOI: 10.1007/978-3-031-94901-2_1
Editora
Springer Nature Switzerland