ML Anthology

Active Planning, Sensing, and Recognition Using a Resource-Constrained Discriminant POMDP

Zhaowen Wang, Zhangyang Wang, Mark Moll, Po-Sen Huang, Devin K. Grady, Nasser M. Nasrabadi, Thomas S. Huang, Lydia E. Kavraki, Mark Hasegawa-Johnson
CVPRW 2014 pp. 754-761
doi:10.1109/CVPRW.2014.116 /cvprw/2014/wang2014cvprw-active/

Abstract

In this paper, we address the problem of object class recognition via observations from actively selected views/modalities/features under limited resource budgets. A Partially Observable Markov Decision Process (POMDP) is employed to find optimal sensing and recognition actions with the goal of long-term classification accuracy. Hetero-geneous resource constraints – such as motion, number of measurements and bandwidth – are explicitly modeled in the state variable, and a prohibitively high penalty is used to prevent the violation of any resource constraint. To improve recognition performance, we further incorporate discrim-inative classification models with POMDP, and customize the reward function and observation model correspond-ingly. The proposed model is validated on several data sets for multi-view, multi-modal vehicle classification and multi-view face recognition, and demonstrates improvement in both recognition and resource management over greedy methods and previous POMDP formulations. 1.

PDF CVPRW Semantic Scholar

Cite

Text

Wang et al. "Active Planning, Sensing, and Recognition Using a Resource-Constrained Discriminant POMDP." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2014. doi:10.1109/CVPRW.2014.116

Markdown

[Wang et al. "Active Planning, Sensing, and Recognition Using a Resource-Constrained Discriminant POMDP." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2014.](https://mlanthology.org/cvprw/2014/wang2014cvprw-active/) doi:10.1109/CVPRW.2014.116

BibTeX

@inproceedings{wang2014cvprw-active,
  title     = {{Active Planning, Sensing, and Recognition Using a Resource-Constrained Discriminant POMDP}},
  author    = {Wang, Zhaowen and Wang, Zhangyang and Moll, Mark and Huang, Po-Sen and Grady, Devin K. and Nasrabadi, Nasser M. and Huang, Thomas S. and Kavraki, Lydia E. and Hasegawa-Johnson, Mark},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2014},
  pages     = {754-761},
  doi       = {10.1109/CVPRW.2014.116},
  url       = {https://mlanthology.org/cvprw/2014/wang2014cvprw-active/}
}