Group-level Speech Emotion Recognition Utilising Deep Spectrum Features
Authors:
Open Access
- 21 October 2020
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 821-826
Abstract
The objectives of this challenge paper are two fold: first, we apply a range of neural network based transfer learning approaches to cope with the data scarcity in the field of speech emotion recognition, and second, we fuse the obtained representations and predictions in a nearly and late fusion strategy to check the complementarity of the applied networks. In particular, we use our Deep Spectrum system to extract deep feature representations from the audio content of the 2020 EmotiW group level emotion prediction challenge data. We evaluate a total of ten ImageNet pre-trained Convolutional Neural Networks, including AlexNet, VGG16, VGG19 and three DenseNet variants as audio feature extractors. We compare their performance to the ComParE feature set used in the challenge baseline, employing simple logistic regression models trained with Stochastic Gradient Descent as classifiers. With the help of late fusion, our approach improves the performance on the test set from 47.88 % to 62.70 % accuracy.Keywords
Funding Information
- Zentrales Innovationsprogramm Mittelstand (ZIM) (16KN069455)
- Deutsche Forschungsgemeinschaft (DFG) (421613952)
- BMW AG
- Group on Language, Audio & Music (GLAM) at Imperial College London
This publication has 23 references indexed in Scilit:
- EmotiW 2016: video and group-level emotion recognition challenges2016
- Face attribute prediction using off-the-shelf CNN features2016
- Deep Residual Learning for Image Recognition2016
- A Review and Meta-Analysis of Multimodal Affect Detection SystemsACM Computing Surveys, 2015
- Automatic Group Happiness Intensity AnalysisIEEE Transactions on Affective Computing, 2015
- CNN Features Off-the-Shelf: An Astounding Baseline for Recognition2014
- Opensmile2010
- Matplotlib: A 2D Graphics EnvironmentComputing in Science & Engineering, 2007
- Solving large scale linear prediction problems using stochastic gradient descent algorithms2004
- A Scale for the Measurement of the Psychological Magnitude PitchThe Journal of the Acoustical Society of America, 1937