We present a series of two studies conducted to understand user’s affective states during voice-based human-machine interactions. Emphasis is placed on the cases of communication errors or failures.
Human behavior modeling publications
Towards Explainable Student Group Collaboration Assessment Models Using Temporal Representations of Individual Student Role and Behavioral Cues
Abstract Collaboration is identified as a required and necessary skill for students to be successful in the fields of Science, Technology, Engineering and Mathematics (STEM). However, due to growing student population and limited teaching staff it is difficult for teachers to provide constructive feedback and instill collaborative skills using instructional methods. Development of simple and […]
Abstract We present a novel generative model for human motion modeling using Generative Adversarial Networks (GANs). We formulate the GAN discriminator using dense validation at each time-scale and perturb the discriminator input to make it translation invariant. Our model is capable of motion generation and completion. We show through our evaluations the resiliency to noise, […]
SRI Authors: Ian M. Colrain
A Hierarchical Behavior Analysis Approach for Automated Trainee Performance Evaluation in Training Ranges
In this paper we present a closed loop mixed reality training system that provides automatic assessment of trainee performance during kinetic military exercises. At the core of our system is a hierarchical behavior analysis approach that integrates a number of data sensor modalities including Audio/Video, RFID and IMUs to automatically capture trainee actions in a comprehensive manner. Our behavior analysis and performance evaluation framework uses a finite state machine (FSM) model in which trainee behaviors are the states of the training scenario and the transitions of states are caused by stimuli that we refer to as trigger events. The goal of behavior analysis is to estimate the states of the trainees with respect to the training scenario and quantify trainee performance. To robustly detect each state, we build classifiers for each behavioral state and trigger event. At a given time, based on the state estimation, a set of related classifiers are activated for detecting trigger events and states that can be transitioned to and from the current states. The overall structure of the FSM and trigger events is determined by a Training Ontology that is specific to the training scenario.
We present a novel approach for multi-modal affect analysis in human interactions that is capable of integrating data from multiple modalities while also taking into account temporal dynamics. Our fusion approach, Joint Hidden Conditional Random Fields (JHRCFs), combines the advantages of purely feature level (early fusion) fusion approaches with late fusion (CRFs on individual modalities) to simultaneously learn the correlations between features from multiple modalities as well as their temporal dynamics. Our approach addresses major shortcomings of other fusion approaches such as the domination of other modalities by a single modality with early fusion and the loss of cross-modal information with late fusion. Extensive results on the AVEC 2011 dataset show that we outperform the state-of-the-art on the Audio Sub-Challenge, while achieving competitive performance on the Video Sub-Challenge and the Audiovisual Sub-Challenge.
SRI Authors: Ajay Divakaran