Strength modelling for real-worldautomatic continuous affect recognition from audiovisual signals

作者：

Highlights：

• Proposed an algorithm of Strength Modelling (SM) to take advantage of various models

• Investigated the effectiveness of SM for value- and time-continuous affect regression

• Evaluated the robustness of SM by integrating it with early and late fusion methods

• The obtained results are comparable to or even better than the related baselines.

摘要

•Proposed an algorithm of Strength Modelling (SM) to take advantage of various models•Investigated the effectiveness of SM for value- and time-continuous affect regression•Evaluated the robustness of SM by integrating it with early and late fusion methods•The obtained results are comparable to or even better than the related baselines.

论文关键词：Strength modelling,Support vector regression,Memory-enhanced recurrent neural networks,Audiovisual affective computing

论文评审过程：Received 24 June 2016, Revised 22 October 2016, Accepted 28 November 2016, Available online 7 December 2016, Version of Record 18 September 2017.

论文官网地址：https://doi.org/10.1016/j.imavis.2016.11.020