Percorrer por autor "Freitas, D."
A mostrar 1 - 3 de 3
Resultados por página
Opções de ordenação
- Automatic speech recognition for Portuguese: a comparative studyPublication . Borghi, P.H.; Teixeira, João Paulo; Freitas, D.
- Editorial: advances in machine learning approaches and technologies for supporting nervous system disease diagnosisPublication . Rodrigues, Pedro; Bispo, Bruno; Freitas, D.; Marques, João; Teixeira, João PauloThe nervous system is essential for physical and mental health but is complex and delicate. As it can unfortunately be affected by several progressive diseases, an early diagnosis is often critical for effective treatment (Xu et al., 2022). The diagnosis of nervous system diseases traditionally relies on a combination of clinical examination, imaging and signals tests, and laboratory tests (Siuly and Zhang, 2016). However, these methods can be time-consuming, expensive, and not always accurate (Milligan, 2019). In an era marked by unprecedented technological advances in machine learning (ML), a computational tool that allows the identification of patterns in data that would be difficult or even impossible for humans, its application to assist in medical diagnosis emerges as a beacon of hope in the complex panorama of nervous system diseases. The Research Topic Advances in machine learning approaches and technologies for supporting nervous system disease diagnosis aims to shed light on the transformative role that ML-based approaches and technologies are playing in reshaping the way an ensemble of nervous system disorders are understood, diagnosed, and treated.
- Nonperiodic pathologic voice signals classification using mel-spectrogram and VGGishPublication . Fernandes, Joana; Pinto, João; Moura, Carla; Vilarinho, Helena; Teixeira, Felipe; Freitas, D.; Teixeira, João PauloIn 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%.