White Paper

June 10, 2015 White Paper

Research Paper on Emotion Recognition

Article Photo

Stream4s in partnership with the University of Surrey in the UK developed a research paper on the use of deep learning methods to further develop emotion recognition technologies.

The paper first reviews the history of automated analysis of human emotions via visual cues. It then explains how deep learning methods make it possible to conduct emotion recognition with raw data, rather than data which has been specially prepared. For this reason it has become the favoured machine-learning method for such analysis.

Further, the research paper explores practical aspects of using deep learning methods for emotion analysis: hardware and system requirements, choice of framework, issues of architecture, data sets, input and output, etc. A number of proposals are offered for extending the analysis of individual images to the realm of video, in order to track how emotions change over time.

“We’re happy to partner with the leading UK university on such an exciting topic as emotion recognition,” said Stream4s CEO Agnis Stelingis. “While other machine-learning methods are available, we’re going to focus on deep learning methods for our research in this area.”

If you are interested in emotion recognition technologies, please do not hesitate to contact us.

Arrow right Back
Office in London, UK
Office in Kaunas, LT
Office in London, UK
Office in Kaunas, LT

Office in Kaunas, LT

Donelaičio g. 62, Kaunas, 50000

Close

Office in London, UK

45 King William Street, London, EC4R 9AN

Close