会议信息

我将在2017年5月12-14日在河南新乡举办的全国智能信息处理学术会议(NCIIP2017)和4月21日在厦门举办的视觉与学习研讨会(VALSE2017)的FCS论坛上汇报我们小组在近五年误差建模方面取得的一些研究成果和探索性认识,欢迎大家参加并给出指导意见!

相关文章见:

[1] Yang Chen, Xiangyong Cao, Qian Zhao, Deyu Meng, Zongben Xu. Denoising Hyperspectral Image with Non-i.i.d. Noise Structure. arXiv:1702.00098, 2017. 

[2]Xiangyong Cao, Qian Zhao, Deyu Meng, Yang Chen, Zongben Xu. Robust Low-rank Matrix Factorization under General Mixture Noise Distributions, IEEE Transactions on Image Processing, 2016.

[3] Xi’ai Chen, Zhi Han, Yao Wang, Qian Zhao, Deyu Meng and Yandong Tang. Robust tensor factorization with unknown noise. CVPR, 2016

[4] Xiangyong Cao, Yang Chen, Qian Zhao, Deyu Meng, Yao Wang, Dong Wang, Zongben Xu. Low-rank Matrix Factorization under General Mixture Noise Distributions. ICCV (oral), 2015.[supplementary material] [Matlab code] [arxiv version]

[5] Qian Zhao, Deyu Meng, Zongben Xu, Wangmeng Zuo, Lei Zhang. Robust principal component analysis with complex noiseSupplementary MaterialICML, 2014.

[6] Deyu Meng, Fernando De la Torre. Robust Matrix Factorization with Unknown Noise. ICCV, 2013.Matlab code.

[7] Qian Zhao, Deyu Meng, Zongben Xu, Wangmeng Zuo, Yan Yan. L1-Norm Low-Rank Matrix Factorization by Variational Bayesian Method. IEEE Transactions on Neural Networks and Learning Systems. 2015. 

[8] Chenqiang Gao, Deyu Meng, Yi Yang, Yongtao Wang, Xiaofang Gao, Alexander G. Hauptmann. Infrared Patch-image Model for Small Target Detection in A Single Image. IEEE Transactions on Image Processing. 22(12): 4996-5009, 2013. Matlab Code 

[9] Deyu Meng, Zongben Xu, Lei Zhang, Ji Zhao. A cyclic weighted median method for L1 low-rank matrix factorization with missing entries. AAAI 2013.

[10] Deyu Meng, Qian Zhao, Zongben Xu. Improve robustness of sparse PCA by L1-norm maximization. Pattern Recognition. 2012, 45(1) 487-497.

 

误差建模的基本原理可通过以下科普文章简要了解:

[11] 孟德宇,误差建模原理,人工智能学会通讯,2016.

 

 

Related Demos for OMoGMF

(Please be patient to wait a few minutes. Some of the gif files need a bit time to be fully depicted.)

 

Related Demos on our submission "Robust Online Dynamic Background Subtraction":

 

1. Experimental results on Li dataset: 

(1) Airport sequence:

 

 

(2) Bootstrap sequence:

 

(3) Curtain sequence:

 

(4) Escalator sequence:

 

(5) Fountain sequence:

 

(6) Shoppingmall sequence:

 

(7) Lobby sequence:

 

(8) Campus sequence:

 

(9) Watersurface sequence:

 

 

t-OMoGMF experients

2. Experimental results on transformed Li dataset:

(1) Airport sequence:

   

 

(2) Curtain sequence:

 

(3) Escalator sequence:

  

 

(4) Fountain sequence:

 

 

(5) Shoppingmall sequence:

  

 

(6) Lobby sequence:

 

 

(7) Campus sequence:

  

 

(8) Watersurface sequence:

 

3. Experimental results on real-world dataset: 

(1) Sidewalk sequence:

 

  

(2) Boulevard sequence:

 

 

(3) Badminton sequence:

 

 

(4) Traffic sequence:

 

4. Performance comparison on real-world dataset:

(1) Sidewalk sequence:

 

 

(2) Boulevard sequence:

 

 

(3) Badminton sequence:

 

(4) Traffic sequence: