Automated Screening of Job Candidate Based on Multimodal Video Processing

Jelena Gorbova, Iiris Lusi, Andre Litvin, Gholamreza Anbarjafari; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2017, pp. 29-35

Abstract


The selection of adequate job candidates is very long and challenging process for each employer. The system presented in this paper is aiming to decrease the time for candidate selection on the pre-employment stage using automatic personality screening based on visual, audio and lexical cues from short video-clips. The system is build to predict candidate scores of 5 Big Personality Traits and to estimate a final decision, to which degree the person from video-clip has to be invited to the job interview. For each channel a set of relevant features is extracted, which are used to train separately from each other using Deep Learning. In the final stage all three results are fused together into final scores prediction. The experiment was conducted on first impression database and achieved significant performance.

Related Material


[pdf]
[bibtex]
@InProceedings{Gorbova_2017_CVPR_Workshops,
author = {Gorbova, Jelena and Lusi, Iiris and Litvin, Andre and Anbarjafari, Gholamreza},
title = {Automated Screening of Job Candidate Based on Multimodal Video Processing},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {July},
year = {2017}
}