Demo for Learning Adaptive Discriminative Correlation Filters (LADCF) via Temporal Consistency preserving Spatial Feature Selection for Robust Visual Tracking

@article{xu2018learning, title={Learning Adaptive Discriminative Correlation Filters via Temporal Consistency Preserving Spatial Feature Selection for Robust Visual Tracking}, author={Xu, Tianyang and Feng, Zhen-Hua and Wu, Xiao-Jun and Kittler, Josef}, journal={arXiv preprint arXiv:1807.11348}, year={2018}}

More powerful features and data augmentation techniques are added for the VOT2018.

Learning Adaptive Discriminative Correlation Filter on Low-dimensional Manifold (LADCF) utilises adaptive spatial regularizer to train low-dimensional discriminative correlation filters. We follow a single-frame learning and updating strategy: the filters are learned after tracking stage and then updated using a fixed rate [1]. We use HOG [2] and CN [3]. Code modules refer to ECO [4] in feature extraction.

• PDollar Toolbox
• mtimesx (https://github.com/martin-danelljan/ECO/tree/master/external_libs/mtimesx)
• mexResize (https://github.com/martin-danelljan/ECO/tree/master/external_libs/mexResize)

Ubuntu 14.04 LTS, Matlab R2016a, CPU Intel(R) Xeon(R) E5-2643

• [1] Henriques, João F., et al. "High-speed tracking with kernelized correlation filters." IEEE Transactions on Pattern Analysis and Machine Intelligence 37.3 (2015): 583-596.
• [2] Dalal, Navneet, and Bill Triggs. "Histograms of oriented gradients for human detection." Computer Vision and Pattern Recognition, 2005. CVPR 2005.
• [3] Van De Weijer, Joost, et al. "Learning color names for real-world applications." IEEE Transactions on Image Processing 18.7 (2009): 1512-1523.
• [4] Danelljan, Martin, et al. "Eco: Efficient convolution operators for tracking." Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017.