Global Pose Estimation with an Attention-Based Recurrent Network

Abstract

The ability for an agent to localize itself within an environment is crucial for many real-world applications. For unknown environments, Simultaneous Localization and Mapping (SLAM) enables incremental and concurrent building of and localizing within a map. We present a new, differentiable architecture, Neural Graph Optimizer, progressing towards a complete neural network solution for SLAM by designing a system composed of a local pose estimation model, a novel pose selection module, and a novel graph optimization process. The entire architecture is trained in an end-to-end fashion, enabling the network to automatically learn domain-specific features relevant to the visual odometry and avoid the involved process of feature engineering. We demonstrate the effectiveness of our system on a simulated 2D maze and the 3D ViZ-Doom environment.

Cite

Text

Parisotto et al. "Global Pose Estimation with an Attention-Based Recurrent Network." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2018. doi:10.1109/CVPRW.2018.00061

Markdown

[Parisotto et al. "Global Pose Estimation with an Attention-Based Recurrent Network." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2018.](https://mlanthology.org/cvprw/2018/parisotto2018cvprw-global/) doi:10.1109/CVPRW.2018.00061

BibTeX

@inproceedings{parisotto2018cvprw-global,
  title     = {{Global Pose Estimation with an Attention-Based Recurrent Network}},
  author    = {Parisotto, Emilio and Chaplot, Devendra Singh and Zhang, Jian and Salakhutdinov, Ruslan},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2018},
  pages     = {237-246},
  doi       = {10.1109/CVPRW.2018.00061},
  url       = {https://mlanthology.org/cvprw/2018/parisotto2018cvprw-global/}
}