Descripción
|
|
---|---|
Cognitive Load (CL) refers to the amount of mental demandthat a given task imposes on an individual?s cognitive systemand it can affect his/her productivity in very high load situa-tions. In this paper, we propose an automatic system capableof classifying the CL level of a speaker by analyzing his/hervoice. Our research on this topic goes into two main directions.In the first one, we focus on the use of Long Short-Term Mem-ory (LSTM) networks with different weighted pooling strate-gies for CL level classification. In the second contribution, forovercoming the need of a large amount of training data, we pro-pose a novel attention mechanism that uses the Kalinli?s audi-tory saliency model. Experiments show that our proposal out-performs significantly both, a baseline system based on SupportVector Machines (SVM) and a LSTM-based system with logis-tic regression attention model | |
Internacional
|
Si |
Nombre congreso
|
Interspeech 2019 |
Tipo de participación
|
960 |
Lugar del congreso
|
Graz- Austria |
Revisores
|
Si |
ISBN o ISSN
|
1990-9772 |
DOI
|
10.21437/Interspeech.2019-1603 |
Fecha inicio congreso
|
15/09/2019 |
Fecha fin congreso
|
19/09/2019 |
Desde la página
|
216 |
Hasta la página
|
220 |
Título de las actas
|
Actas del Congreso Interspeech 2019 |