|International Computer Science and Engineering Society (ICSES)|
ICSES Transactions on Image Processing and Pattern Recognition
Vol. 4, No. 4, Dec. 2018
Deep Emotional Intelligence: Study on Discrete Action Sequences | Book Chapter of Image Segmentation: A Guide to Image Mining
Santhoshkumar R a,, Kalaiselvi Geetha M b
Highlights and Novelties
Automatic emotion recognition is becoming recent research focus today. A facet of human intelligence is the ability to recognize emotion that is regarded as one of the attribute of emotional intelligence. Although research based on facial expressions or speech is seen in thrive, recognizing emotions from body gestures has been remained as a less explored topic. This chapter proposes a machine learning approach to achieve emotional intelligence. A set of Different Bin Level HoG features (DBLHoG) and Spatio-Temporal Interest Points (STIP) are extracted from human body movements present in each frame and are fed to a supervised learning algorithm. This experiment is conducted by GEMEP corpus dataset. In this dataset human expressing the five archetypical emotions likes (anger, joy, sad, fear and pride) using body movements. In this emotions recognition problem, random forest classifier outperformed the kNN classifier by achieving an overall recognition accuracy of 94.8% for DBLHoG feature. Moreover, the performance can be measured by qualitative approach. Finally, this chapter gives a brief study on achieving emotional intelligence with a deep learning approach.
© Copyright was transferred to International Computer Science and Engineering Society (ICSES) by all the Authors.
Cite this manuscript as
Santhoshkumar R, Kalaiselvi Geetha M, "Deep Emotional Intelligence: Study on Discrete Action Sequences," in Image Segmentation: A Guide to Image Mining, 1st ed., ITIPPR: ICSES, 2018, pp. 83-94.
For External Scientific Databeses
Written by: Admin | Link ...