1. Proceedings
  2. I3M
  3. 2018
  4. MAS
  5. 10.46354/i3m.2018.mas.008

Leveraging virtual environments to train a Deep Learning algorithm

  • a Dean Reed , 
  • b Troyle Thomas  , 
  • c Latika Eifert ,
  • d Shane Reynolds ,
  • e Jonathan Hurter ,
  • f  Frank Tucher 
  • a,b,d,e Institute for Simulation and Training, University of Central Florida
  • c,f Army Research Laboratory-Human Research and Engineering Directorate Advanced Training and Simulation Division, SFC Paul Ray Smith Center
Cite as
Reed D., Thomas T., Eifert L., Reynolds S., Hurter J., Tucker F. (2018). Leveraging virtual environments to train a Deep Learning algorithm. Proceedings of the 17th International Conference on Modeling & Applied Simulation (MAS 2018), pp. 48-54. DOI: https://doi.org/10.46354/i3m.2018.mas.008

Abstract

Open source datasets are the typical source used to train computers to accurately detect visual objects (e.g., humans, animals, and inanimate objects) through various machine learning methods (e.g., Deep Learning (DL)). This data, however, is not feasible for use in the military domain. In this paper, a comparative analysis of real and virtual training data is provided, using the You Only Look Once (YOLO) Convolutional Neural Network (CNN) model. The main concern of this paper is to verify the process and accuracy of using a domain-specific U.S. Army Virtual Environment (VE), in contrast to a Real Environment (RE) dataset, with DL. Comparative results suggest that substituting a VE to provide training data for the DL model saves manual labour while maintaining a quality precision-recall curve.

References

  1. Atapour-Abarghouei A. and Breckon T.P., 2018. Realtime monocular depth estimation using synthetic data with domain adaptation. Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). June 18-22, Salt Lake City (Utah, USA).
  2. Baker T., 2018. VBS3 Resource. Milgaming Website. Available from: https://milgaming.army.mil/VBS3/files/ResourceList.aspx [accessed 11 July 2018].
  3. Broggi A., Fascioli A., Grisleri P., Graf T., and Meinecke M., 2005. Model-based validation
    approaches and matching techniques for automotive vision-based pedestrian detection. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Workshops. September 21-23, San Diego (California, USA).
  4. Deng J., Russakovsky O., Krause J., Bernstein M.S., Berg A., and Fei-Fei L., 2014. Scalable multi-label annotation. Proceedings of CHI 2014, One of a CHInd, pp 3099-3102. April 26 – May 1, Toronto (Ontario, Canada).
  5. Everingham M., Eslami S.A., Van Gool L., Williams C.K., Winn J., and Zisserman A., 2015. The pascal visual object classes challenge: A retrospective. International Journal of Computer Vision, 111 (1), 98-136.
  6. Gaidon A., Wang Q., Cabon Y., and Vig E., 2016. Virtual worlds as proxy for multi-object tracking analysis. arXiv. Available from https://arxiv.org/abs/1605.06457 [accessed 4 May 2018].
  7. Gretton A., Smola A., Huang J., Schmittfull M., Borgwardt K., and Schölkopf B., 2009. Covariate shift in kernel mean matching. In: J Quiñonero-Candela., M. Sugiyama, A. Schwaighofer, and N.D Lawrence., eds. Dataset shift in machine learning. Cambridge, MA: MIT, 131-160.
  8. Lee M., Valisetty R., Breuer A., Kirk K., Panneton B., and Brown S., 2018. Current and future applications of machine learning for the US Army.  Aberdeen Proving Ground, MD: US Army Research Laboratory.
  9. Rajpura P.S., Hegde R.S., and Bojinov H., 2017. Object detection using deep CNNs trained on synthetic images. arXiv. Available from: https://arxiv.org/abs/1706.06782 [accessed 4 May 2018].
  10. Redmon J., Divvala S., Girshick R., and Farhadi A., 2016.You Only Look Once: Unified, real-time object detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 779-788. June 27-30, Las Vegas, (Nevada, USA).
  11. System Studies & Simulation, Inc, 2015. GSA Schedule. GSA Advantage! Available from https://www.gsaadvantage.gov/ref_text/GS00F0037P/0P5FUQ.38UNOH_GS-00F-0037P_PSSREV20151222MODS003600380039. PDF [accessed 11 July 2018].
  12. U.S. Army Research Laboratory, 2017. U.S. Army Research Laboratory Essential Research Areas. United States Army Research Laboratory. Available from: https://www.arl.army.mil/www/default.cfm?page=2401 [accessed June 21 2018]