ABSTRACT
As a fundamental technology of human motion analysis, video based pose estimation has attracted more and more attention of researchers. While most mainstream methods treated videos as a collection of unrelated frames, random false positives caused by light and shadow changes will bring greater uncertainty to the results of human body estimation. To eliminate these false positives in multi-person 2D pose estimation, a method of multi-person 2D pose estimation using the time-space relationship of adjacent frames is proposed. At first, Frame Association Information(FAI) is defined to represent the relationship between person objects in adjacent frames, which contents similarity of position, size and poses. Then, a method of false positives eliminating is designed using FAIs between frames, which can identify and remove false positives in frames according to their similarities to human objects in a benchmark frame. The following experiments show the effectiveness of our method.
- M. Andriluka, L. Pishchulin, P. Gehler, B. Schiele, C. 2014. 2D human pose estimation: New benchmark and state of the art analysis. In CVPR 2014(Columbus, OH, USA, June 23-28), 3686--3693. Google ScholarDigital Library
- L. Pishchulin and M. Andriluka, P. Gehler, B. Schiele, C. 2013. Poselet conditioned pictorial structures. In CVPR 2013(Portland, OR, June), 588--595. Google ScholarDigital Library
- L. Pishchulin, M. Andriluka, P. Gehler, and B. Schiele, C. 2013. Strong appearance and expressive spatial models for human pose estimation. In ICCV 2013(Sydney, NSW, Australia, Dec 1-8), 3487--3494. Google ScholarDigital Library
- Y. Yang and D. Ramanan, C. 2011. Articulated pose estimation with flexible mixtures-of-parts. In CVPR 2011(Colorado Springs, CO, USA, USA, June 20-25), 1385--1392. Google ScholarDigital Library
- M. Sun and S. Savarese, C. 2011. Articulated part-based model for joint object detection and pose estimation. In ICCV 2011(Barcelona, Spain, Nov 6-13), 723--730. Google ScholarDigital Library
- Y. Tian, C. L. Zitnick, S. G. Narasimhan, C. 2012. Exploring the spatial hierarchy of mixture models for human pose estimation. In ECCV 2012(Florence, Italy, Oct 7-13), 256--269. Google ScholarDigital Library
- M. Dantone, J. Gall, C. Leistner, and L. Van Gool, C. 2013. Human pose estimation using body parts dependent joint regressors. In CVPR 2013(Portland, OR, USA, June 23-28), 3041--3048. Google ScholarDigital Library
- L. Karlinsky and S. Ullman, C. 2012. Using linking features in learning non-parametric part models. In ECCV 2012(Florence, Italy, Oct 7-13), 326--339. Google ScholarDigital Library
- J. Carreira, P. Agrawal, K. Fragkiadaki, and J. Malik, C. 2016. Human pose estimation with iterative error feedback. In CVPR 2016(Las Vegas, NV, USA, June 27-30), 4733--4742.Google Scholar
- T. Pfister, J. Charles, and A. Zisserman, C. 2015. Flowing convents for human pose estimation in videos. In ECCV 2015(Santiago, Chile, Dec 7-13), 1913--1921. Google ScholarDigital Library
- L. Pishchulin, E. Insafutdinov, S. Tang, B. Andres, M. Andriluka, P. Gehler, and B. Schiele, C. 2016. Deepcut: Joint subset partition and labeling for multi person pose estimation. In CVPR 2016(Las Vegas, NV, USA, June 27-30), 4929--4937.Google Scholar
- J. Tompson, R. Goroshin, A. Jain, Y. LeCun, and C. Bregler, C. 2015. Efficient object localization using convolutional networks. In CVPR 2015(Boston, Massachusetts, USA, June 7-12), 648--656.Google Scholar
- S.E. Wei, V. Ramakrishna, T. Kanade, Y. Sheikh, C. 2016. Convolutional pose machines. In CVPR 2016(Las Vegas, NV, USA, June 27-30), 4724--4732.Google ScholarCross Ref
- A. Newell, K. Yang, and J. Deng, C. 2016. Stacked hourglass networks for human pose estimation. In ECCV 2016(Amsterdam, The Netherlands, Oct 11-14), 483--499.Google Scholar
- E. Insafutdinov, L. Pishchulin, B. Andres, M. Andriluka, and B. Schiele, C. 2016. Deepercut: A deeper, stronger, and faster multi-person pose estimation model. In ECCV 2016(Amsterdam, The Netherlands, Oct 11-14), 34--50.Google Scholar
- Zhe. Cao, T. Simon, S.-E. Wei, and Y. Sheikh, C. 2017. Realtime multi-person 2d pose estimation using part affinity fields. In CVPR 2017(Honolulu, HI, USA, July 21-26), 1302--1310.Google Scholar
- G. Papandreou, T. Zhu, N. Kanazawa, A. Toshev, J. Tompson, C. Bregler, and K. Murphy, C. 2017. Towards accurate multi-person pose estimation in the wild. In CVPR 2017(Honolulu, HI, USA, July 21-26), 3711--3719.Google Scholar
- F. S. Khan, J. van de Weijer, and M. Vanrell. 2012. Modulating shape features by color attention for object recognition. J. International Journal of Computer Vision.(May, 2012), 49--64. Google ScholarDigital Library
Index Terms
False Positive Eliminating Using Frame Association Information
Recommendations
The False Dichotomy between Positive and Negative Affect in Game Play
CHI PLAY '15: Proceedings of the 2015 Annual Symposium on Computer-Human Interaction in PlayMost of the time games make us happy, but sometimes they are frustrating or make us feel sad. They allow us to experience pleasure, success and joy, but they can also yield feelings of frustration, failure, or sorrow from darker themes. In games, we can ...
QuickPose: Real-time Multi-view Multi-person Pose Estimation in Crowded Scenes
SIGGRAPH '22: ACM SIGGRAPH 2022 Conference ProceedingsThis work proposes a real-time algorithm for reconstructing 3D human poses in crowded scenes from multiple calibrated views. The key challenge of this problem is to efficiently match 2D observations across multiple views. Previous methods perform multi-...
Multi-person 3D pose estimation from 3D cloud data using 3D convolutional neural networks
AbstractHuman pose estimation is considered one of the major challenges in the field of Computer Vision, playing an integral role in a large variety of technology domains. While, in the last few years, there has been an increased number of ...
Highlights- 3D-CNN architecture for 3D human pose estimation from 3D/depth data.
- Fully-...
Comments