Abstract Typical Local Decorrelation Channel Feature (LDCF) for pedestrian detection generates filters derived from decorrelation for each entire positive sample, using Principle Component Analysis (PCA) method. Meanwhile, extensive pedestrian detection methods, which utilize statistic human shape to guide filters design, point out that the head-shoulder area is the most discriminative patches in typical classification stage. Inspired by above mentioned local decorrelation operation and discriminative areas that most classifiers indicate, in this paper we propose to integrate human shape priority into image patch decorrelation to generate novel filters. To be specific, we extract covariance from salient patches that contain discriminative features, instead of each entire positive sample. Furthermore, we also propose to share covariance matrix within grouping channels. Our method is efficient as it avoids extracting uninformative filters from redundant covariance of convergent patches, due to embedded prior human shape info. Experiments on INRIA and Caltech-USA public pedestrian dataset has been done to demonstrate effectiveness of our proposed methods. The result shows that our proposed method could decrease log-average miss rate with detection speed retained compared to LDCF and most non-deep methods.

References

[1] Al-Najdawi N., Tedmori S., Edirisinghe E., and Bez H., “An Automated Real-Time People Tracking System Based on KLT Features Detection,” The International Arab Journal of Information Technology, vol. 9, no. 1, pp. 100- 107, 2012.

[2] Baek J., Kim J., and Kim E., “Fast and Efficient Pedestrian Detection via the Cascade Implementation of an Additive Kernel Support Vector Machine,” Transactions on Intelligent Transportation Systems, vol. 18, no. 4, pp. 902- 916, 2017.

[3] Bastian B. and Jiji C., “Aggregated Channel Features with Optimum Parameters for Pedestrian Detection,” in Proceedings of International Conference on Pattern Recognition and Machine Intelligence, Kolkata, pp. 155-161, 2017.

[4] Cao J., Pang Y., and Li X., “Pedestrian Detection Inspired By Appearance Constancy and Shape Symmetry,” IEEE Transactions on Image Processing, vol. 25, no. 12, pp. 5538-5551, 2016.

[5] Dollár P., Tu Z., Perona P., and Belongie S., “Integral Channel Features,” in Proceedings of British Machine Vision Conference, London, pp. 91-98, 2009.

[6] Dollár P., Wojek C., Schiele B., and Perona P., “Pedestrian Detection: An Evaluation of the State of the Art,” Transactions on Pattern Analysis and Machine Intelligence, vol. 34, no. 4, pp. 743-761, 2012.

[7] Du X., El-Khamy M., Lee J., and Davis L., “Fused Dnn: A Deep Neural Network Fusion Approach to Fast and Robust Pedestrian Detection,” in Proceedings of IEEE Winter Conference on Applications of Computer Vision, Santa Rosa, pp. 953-961, 2017.

[8] Lim J., Zitnick C., and Dollár P., “Sketch Tokens: A Learned Mid-Level Representation for Contour and Object Detection,” in Proceedings/CVPR, IEEE Computer Society Conference on Computer Vision and Pattern Recognition IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Portland, pp. 3158-3165, 2013.

[9] Nam W., Dollár P., and Han J., “Local Decorrelation for Improved Detection,” in Proceedings of the 27th International Conference on Neural Information Processing Systems, Montréal, pp. 424-432, 2014.

[10] Ohn-Bar E. and Trivedi M., “To Boost or Not to Boost? on The Limits of Boosted Trees for Object Detection,” in Proceedings of International Conference on Pattern Recognition, Cancun, pp. 3350-3355, 2016.

[11] Ouyang W., Zhou H., Li H., Li Q., Yan J., and Wang X., “Jointly Learning Deep Features, Deformable Parts, Occlusion and Classification for Pedestrian Detection,” Transactions on Pattern Analysis and Machine Intelligence, vol. 40, no. 8, pp. 1874-1887, 2017.

[12] Shen J., Zuo X., Li J., Yang W., and Ling H., “A Novel Pixel Neighborhood Differential Statistic Feature for Pedestrian and Face Detection,” Pattern Recognition, vol. 63, pp. 127-138, 2017.

[13] Shen J., Zuo X., Yang W., Yu H., and Liu G., “Learning Discriminative Shape Statistics Distribution Features for Pedestrian Detection,” Neurocomputing, vol. 184, pp. 66-77, 2016.

[14] Tian Z., Shen C., Chen H., and He T., “FCOS: Fully Convolutional One-Stage Object Detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, pp. 9627-9636, 2019.

[15] Viola P. and Jones M., “Robust Real-Time Face Detection,” International Journal of Computer Vision, vol. 57, no. 2, pp.137-154, 2004.

[16] Zhang L., Lin l., Liang X., and He K., “Is Faster R-Cnn Doing Well for Pedestrian Detection?,” in Proceedings of European Conference on Computer Vision, Amsterdam, pp. 443-457, 2016.

[17] Zhang S., Bauckhage C., and Cremers A., “Efficient Pedestrian Detection via Rectangular Features Based on a Statistical Shape Model,” IEEE Transactions on Intelligent Transportation Systems, vol. 16, no. 2, pp. 763-775, 2014.

[18] Zhang S., Benenson R., and Schiele B., “Filtered Channel Features for Pedestrian Detection,” Computer Vision and Pattern Recognition, vol. 1, no. 2, pp. 1751-1760, 2015.

[19] Zhang S., Benenson R., Omran M., Hosang J., and Schiele B., “How Far are we from Solving Pedestrian Detection?” in Proceedings of Computer Vision and Pattern Recognition, Las Vegas, pp. 1259-1267, 2016.

[20] Zhang S., Benenson R., Omran M., Hosang J., and Schiele B., “Towards Reaching Human Performance in Pedestrian Detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 40, no. 4, pp. 973-986, 2017.

[21] Zhao Y., Yuan Z., Chen D., Lyu J., and Liu T., “Fast Pedestrian Detection via Random Projection Features with Shape Prior,” in Proceedings of IEEE Winter Conference on Applications of Computer Vision, Santa Rosa, pp. 962-970, 2017.

[22] Zhou X., Wang D., and Krähenbühl P., “Objects as Points,” in Proceedings of Computer Vision and Pattern Recognition, Long Beach, pp. 141- 183, 2019. Specific Patches Decorrelation Channel Feature on Pedestrian Detection 503 Xue-ming Ding received the Ph.D. Degree in control science and engineering from University of Science and Technology of China in 2005. He is currently an associate professor with University of Shanghai for Science and Technology. His research interests include system identification, embedded system and machine learning. Dong-fei Ji received his first degree and M.S. degree both from the University of Shanghai for Science and Technology, Shanghai, China, all in control science and engineering. His interested topics are optical character recognition, pedestrian detection, pedestrian ReID and machine learning.