The second keynote in AffectiCom'21 will be presented by Ajjen Joshi.
Title Computational Human Sensing: Applications of Face, Gesture and Affect Analysis and Synthesis
Abstract Computational analyses of human face and gesture signals encompass a myriad of challenging research problems involving computer vision, machine learning and human computer interaction. In this talk, I will present work on the following challenges: a) accurately estimating human eye-gaze signals for inferring visual attention, and developing fast, hands-free, eye-gaze-based text-entry methods; b) classifying hand and body gestures along with the temporal localization of their occurrence in a continuous stream, with a focus on personalization; c) recognizing emotional and cognitive states from faces, with applications ranging from detecting driver drowsiness and frustration to facial expressivity levels in people with Parkinson’s Disease; and d) synthesizing data in input as well as feature space in order to alleviate the challenges of data paucity and class imbalance. I will discuss contemporary research challenges in the field of affective computing and possible future directions in this exciting realm.
Bio Ajjen Joshi is a Senior Research Scientist at Affectiva, where he works on various problems in human perception and affective computing. He finished his PhD from Boston University in 2018, where he was a member of the Image and Video Computing research group. He graduated with a BA from Connecticut College in 2012, where he majored in Computer Science and Architectural Studies.