Identification of emergent leaders in a meeting scenario using multiple kernel learning
read more
Citations
Prediction of the Leadership Style of an Emergent Leader Using Audio and Visual Nonverbal Features
Robust eye contact detection in natural multi-person interactions using gaze and speaking behaviour
Understanding nonverbal communication cues of human personality traits in human-robot interaction
Investigation of Small Group Social Interactions Using Deep Visual Activity-Based Nonverbal Features
Moving as a Leader: Detecting Emergent Leadership in Small Groups using Body Pose
References
Rapid object detection using a boosted cascade of simple features
SMOTE: synthetic minority over-sampling technique
SMOTE: Synthetic Minority Over-sampling Technique
A comparison of methods for multiclass support vector machines
Performance of optical flow techniques
Related Papers (5)
Frequently Asked Questions (14)
Q2. What are the future works in "Identification of emergent leaders in a meeting scenario using multiple kernel learning" ?
As future work, audio nonverbal features will be extracted and combined with the existing visual nonverbal features, assuming that their combination might produce much better results.
Q3. What features were the performing to infer the ELs?
Speaking turn based features, body activity based features and energy together were the best performing feature combination to infer the ELs.
Q4. What is the way to infer head activity?
Instead of using the optical flow vectors, using the absolute displacement of center of face bounding boxes in consecutive frames can be an alternative way to infer the head motion.
Q5. What is the way to combine different kernels?
The simplest way to combine different kernels is to use an un-weighted sum of kernel function which gives equal preferences to all kernels.
Q6. What is the performing method when all VFOA based features were used?
The best performing method when all VFOA based features were used was SVM-cost [9] for the most EL detection and the best performing method to detect the least EL was RLFA.
Q7. What is the significance of the kernel weights obtained from LMKL?
The kernel weights obtained from LMKL can be used to extract the relative contributions of features when all features are concatenated [11, 12].
Q8. What are the first features used for emergent leadership?
for the first time, VFOA, head activity and body activity based features are used together for emergent leadership.
Q9. How many independent judges were used to observe each meeting?
In detail, two independent judges were used to observe each meeting and rate each participant using SMYLOG (called as SYMLOG-Observers in this paper) and GLIS (called GLIS-Observers in this paper).
Q10. How many observers annotated each meeting segment?
Each observers annotated either 12 or 13 meeting segments in total, while no more than one segment which belongs to the same meeting session was annotated by the same observer.
Q11. What is the main reason of the poor performance of nonverbal visual features?
The main reason of poor performance of nonverbal visual features can be the insufficient performance of the method used to extract VFOA automatically (42% frame level accuracy on a subsample of the data).
Q12. What are the common nonverbal features used in the study?
It is also detected that the majority of the nonverbal features used werehighly correlated with the results of the social psychology questionnaires which test the leadership and the dominance.
Q13. What are the common tasks in small group decision making, dominance and leadership?
The participants performed either “winter survival” or “desert survival” [16] tasks as these are the most common tasks in small group decision making, dominance and leadership.
Q14. What are the nonverbal visual features extracted from the annotations?
These nonverbal visual features include i) the visual focus of attention (VFOA) based features which were extracted using the estimation of the head pose, ii) the head activity based features which were extracted using face detection and optical flow, and iii) the body activity based features that were obtained using image differencing.