The Frontiers of Society, Science and Technology, 2020, 2(9); doi: 10.25236/FSST.2020.020909.
Cheng Huaihui1, Zhang Shengbo2
1 Nanyang Technological University, Singapore 639798
2 Carnegie Mellon University, Pittsburgh, PA 15213, United States
In order to solve the problem of low tracking accuracy caused by the lack of appearance model and motion model in multi instance tracking algorithm, this paper proposes a multi instance deep learning target tracking algorithm. In the original multi example tracking algorithm, the motion model only assumes that the motion of the object between frames will not exceed a certain range, which can not effectively reflect the motion state of the target. Particle filter algorithm is introduced to predict the target and improve the accuracy of tracking. The experimental results of different image sequences in complex environment show that the proposed algorithm has higher tracking accuracy and better robustness than other tracking algorithms.
Target tracking, Multi example learning, Deep learning
Cheng Huaihui, Zhang Shengbo. Multi Instance Deep Learning Target Tracking. The Frontiers of Society, Science and Technology (2020) Vol. 2 Issue 9: 37-39. https://doi.org/10.25236/FSST.2020.020909.
[1] Wu Y, Lim J, and Yang M H. Online object tracking: a benchmark[C]. Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, Portland, USA, 2013: 2411-2418.
[2] Ross D A, Lim J, Lin R S, et al (2008). Incremental learning forrobust visual tracking [J]. International Journal of Computer Vision, vol.77, no.3, pp.125-141.
[3] Zhang K, Zhang L, Yang M H (2014). Fast compressive tracking [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.36, no.10, pp.2002-2015.
[4] Zhong W, Lu H C, Yang M H (2014). Robust object tracking via sparse collaborative appearance model [J]. IEEE Transactions on Image Processing, vol.23, no.5, pp.2356-2368.
[5] Zhang K, Zhang L, Liu Q, et al (2014). Fast tracking via densespatio-temporal context learning[C]. Proceedings of European Conference on Computer Vision, Zurich, Switzerland, pp.127-141.
[6] Kalal Z, Mikolajczyk K, Matas J (2012). Tracking-learning-detection [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.34, no7, pp.1409-1422.
[7] Babenko B, Yang M H, Belongie S (2011). Robust object tracking with online multiple instance learning [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.33, no.8, pp.1619-1632.
[8] Zhang K H, Song H H (2013). Real-time visual tracking via online weighted multiple instance learning [J]. Pattern Recognition, vol.46, no.1, pp.397-411.