Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 3
Encontrada(s) 1 página(s)

UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

LSTM neural network-based speaker segmentation using acoustic and language modelling

  • India Massana, Miquel Àngel|||0000-0002-3107-3662
  • Rodríguez Fonollosa, José Adrián|||0000-0001-9513-7939
  • Hernando Pericás, Francisco Javier|||0000-0002-1730-8154
This paper presents a new speaker change detection system based on Long Short-Term Memory (LSTM) neural networks using acoustic data and linguistic content. Language modelling is combined with two different Joint Factor Analysis (JFA) acoustic approaches: i-vectors and speaker factors. Both of them are compared with a baseline algorithm that uses cosine distance to detect speaker turn changes. LSTM neural networks with both linguistic and acoustic features have been able to produce a robust speaker segmentation. The experimental results show that our proposal clearly outperforms the baseline system., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

LSTM neural network-based speaker segmentation using acoustic and language modelling

  • India Massana, Miquel Àngel
  • Rodríguez Fonollosa, José Adrián
  • Hernando Pericás, Francisco Javier
This paper presents a new speaker change detection system based on Long Short-Term Memory (LSTM) neural networks using acoustic data and linguistic content. Language modelling is combined with two different Joint Factor Analysis (JFA) acoustic approaches: i-vectors and speaker factors. Both of them are compared with a baseline algorithm that uses cosine distance to detect speaker turn changes. LSTM neural networks with both linguistic and acoustic features have been able to produce a robust speaker segmentation. The experimental results show that our proposal clearly outperforms the baseline system., Peer Reviewed, Postprint (published version)
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Contribución de congreso (ConferenceOutput)

Classifying depression symptom severity: assessment of speech representations in personalized and generalized machine learning models

  • Campbell Hernández, Edward Lázaro
  • Dineley, Judith
  • Conde, Pauline
  • Matcham, Faith
  • White, Katie M.
  • Oetzmann, Carolin
  • Simblett, Sara
  • Bruce, Stuart
  • Folarin, Amos A.
  • Wykes, Til
  • Vairavan, Srinivasan
  • Dobson, Richard J. B.
  • Docío Fernández, Laura
  • García Mateo, Carmen
  • Narayan, Vaibhav A.
  • Hotopf, Matthew
  • Cummins, Nicholas
There is an urgent need for new methods that improve the management and treatment of Major Depressive Disorder (MDD). Speech has long been regarded as a promising digital marker in this regard, with many works highlighting that speech changes associated with MDD can be captured through machine learning models. Typically, findings are based on cross-sectional data, with little work exploring the advantages of personalization in building more robust and reliable models. This work assesses the strengths of different combinations of speech representations and machine learning models, in personalized and generalized settings in a two-class depression severity classification paradigm. Key results on a longitudinal dataset highlight the benefits of personalization. Our strongest performing model set-up utilized self-supervised learning features and convolutional neural network (CNN) and long short-term memory (LSTM) back-end., National Institute for Health and Care Research (NIHR) Maudsley Biomedical Research Centre (BRC), South London and Maudsley NHS Foundation Trust, UK Medical Research Council | Ref. MR/N013700/1, King’s College London, Xunta de Galicia | Ref. ED431B 2021/24, Fondo Europeo de Desarrollo Regional