Multi-domain and multi-task prediction of extraversion and leadership from meeting videos

Abstract

Automatic prediction of personalities from meeting videos is a classical machine learning problem. Psychologists define personality traits as uncorrelated long-term characteristics of human beings. However, human annotations of personality traits introduce cultural and cognitive bias. In this study, we present methods to automatically predict emergent leadership and personality traits in the group meeting videos of the Emergent LEAdership corpus. Prediction of extraversion has attracted the attention of psychologists as it is able to explain a wide range of behaviors, predict performance, and assess risk. Prediction of emergent leadership, on the other hand, is of great importance for the business community. Therefore, we focus on the prediction of extraversion and leadership since these traits are also strongly manifested in a meeting scenario through the extracted features. We use feature analysis and multi-task learning methods in conjunction with the non-verbal features and crowd-sourced annotations from the Video bLOG (VLOG) corpus to perform a multi-domain and multi-task prediction of personality traits. Our results indicate that multi-task learning methods using 10 personality annotations as tasks and with a transfer from two different datasets from different domains improve the overall recognition performance. Preventing negative transfer by using a forward task selection scheme yields the best recognition results with 74.5% accuracy in leadership and 81.3% accuracy in extraversion traits. These results demonstrate the presence of annotation bias as well as the benefit of transferring information from weakly similar domains.

Recent trends in deep learning based personality detection

“…Multimodal [38] Facial features extracted by VGG-Face, Scene features by VGG-VD19 and combined with audio features using Random Forest based score level fusion [56] Multi task learning of learning leadership and extraversion simultaneously and perceived personality are in-fact quite similar. One of the most important and widely used dataset in the field of multimodal perceived personality recognition is the ChaLearn First Impressions dataset.…”

“…Visual [86] Histogram of Oriented Gradients (HOG) for modeling the face, EigenFace and specific points on the face were taken and fed into SVM, Gradient Boosting, bTree [5] CERT used for extracting facial expressions followed by thresholding as well as Hidden Markov Models [39] Transfer Learning, pre-trained VGG-Face and VGG-19 to extract facial as well as background features followed by regularized regression with a kernel ELM classifier [2] Polynomial and Radial Basis Function (RBF) kernels for SVMs Multimodal [38] Facial features extracted by VGG-Face, Scene features by VGG-VD19 and combined with audio features using Random Forest based score level fusion [56] Multi task learning of learning leadership and extraversion simultaneously Fig. 3 Segment-level occlusion analysis.…”

Emergent Leadership Detection Across Datasets

“…Work on ELEA investigated emergent leadership detection from recordings of the meetings, by using audio-and visual or multi-modal features [24,25], and more recently by using features obtained from a co-occurrence mining procedure [21]. Kindiroglu et al investigated domain adaptation and multi-task learning in order to predict leadership and extraversion on ELEA using video blogs annotated with personality impressions [18]. Their work is different to the cross-dataset setting described above, as they assumed access to emergent leadership ground truth on ELEA.…”

Towards a model of nonverbal leadership in unstructured joint physical activity

“…The ability to determine who leads a group of interacting people is an important research topic both in Computer and Social Sciences. Several computational approaches were proposed to automatically determine the leadership [1,8,19,22]. Most of them focus on human-human interactions in which the verbal communication is complemented by some nonverbal behaviors.…”

“…Most of them focus on human-human interactions in which the verbal communication is complemented by some nonverbal behaviors. A typical example is the meetings (e.g., [8,19]). These papers exploit nonverbal cues of leadership that are related to the main activity (i.e., speaking and listening), such as the variations of prosody, the number of turntaking, and so on.…”