Visual focus of attention estimation based on improved hybrid incremental dynamic Bayesian network

Abstract

In this paper, a visual focus of attention (VFOA) detection method based on the improved hybrid incremental dynamic Bayesian network (IHIDBN) constructed with the fusion of head, gaze and prediction sub-models is proposed aiming at solving the problem of the complexity and uncertainty in dynamic scenes. Firstly, gaze detection sub-model is improved based on the traditional human eye model to enhance the recognition rate and robustness for different subjects which are detected. Secondly, the related sub-models are described, and conditional probability is used to establish regression models respectively. Also an incremental learning method is used to dynamically update the parameters to improve adaptability of this model. The method has been evaluated on two public datasets and daily experiments. The results show that the method proposed in this paper can effectively estimate VFOA from user, and it is robust to the free deflection of the head and distance change.

References

[1] DasD, HoqueM M, OnukiT. The 21st IEEE Interna tional Symposium on Robot and Human Interactive Communication, 2012, 29: 496

[2] DasD, KobayashiY, KunoY. Attracting Attention and Establishing a Communication Channel Based on the Level of Visual Focus of Attention, IEEE/RSJ International Conference on Intelligent Robots & Systems, 2013, Tokyo, Japan, IEEE Press: 2149

[3] DasD, RashedM G, KobayashiY. IEEE Transactions on Human-Machine Systems, 2015, 45: 664

[4] SheikhiS, OdobezJ M. Pattern Recognition Letters, 2014, 66: 81

[5] LanillosP, FerreiraJ F, DiasJ. International Journal of Approximate Reasoning, 2017, 87: 1

[6] NaqviR A, ArsalanM, BatchuluunG. Sensors, 2018, 18: 456

[7] WangH, AntonelliM. Using Point Cloud Data to Improve Three Dimensional Gaze Estimation, IEEE International Conference on Engineering in Medicine and Biology Society, Jeju Island, 2017, Korea, IEEE Press: 795

[8] TawariA, MogelmoseA, MartinS. Attention Estimation by Simultaneous Analysis of Viewer and View, IEEE International Conference on Intelligent Transportation Systems, 2010, Funchal, Portugal, IEEE Press: 1381

[9] LuF, SuganoY, OkabeT. IEEE Transactions on Image Processing, 2015, 24: 3680

[10] Hua-biaoQ, Xin-liangW, JieL. Electronic Journal, 2016, 44: 420

[11] Xin-chanZ, Shi-haoP, Ya-pingW. Journal of System Simulation, 2018, 30: 36

[12] ViolaP, JonesM. Rapid Object Detection using a Boosted Cascade of Simple Features, IEEE International Conference on Computer Vision & Pattern Recognition. IEEE Computer Society, 2001, 511

[13] Ti-zhouQ, Shu-lingD. Journal of Computer Aided Design and Graphics, 2014, 25: 1151

[14] Cheng-longL, FanZ, XingM. Journal of Computer Aided Design and Graphics, 2017, 29: 2309

[15] ChenJ, JiQ. IEEE Transactions on Image Processing, 2015, 24: 1076

[16] ShmelkovK, SchmidC, AlahariK. Incremental Learning of Object Detectors without Catastrophic Forgetting, IEEE International Conference on Computer Vision (ICCV), Venice, 2017, Italy, IEEE Computer Society: 3420

[17] GanjewarP D, BaraniS, WaghS J. Data Reduction Using Incremental Naive Bayes Prediction (INBP) in WSN, International Conference on Information Processing, 2016, Kyoto, Japan, IEEE Press: 398

[18] ZhangX, SuganoY, FritzM. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 41: 162

[19] WangY, ShenT, YuanG. Knowledge-Based Systems, 2016, 110: 293

[20] YeN, TaoX, DongL. Mouse Calibration Aided Real-time Gaze Estimation Based on Boost Gaussian Bayesian learning, IEEE International Conference on Image Processing (ICIP), 2016, Phoenix, Arizona, USA, IEEE Press: 2797