Dense Optical Flow Prediction from a Static Image

@article{Walker2015DenseOF,
  title={Dense Optical Flow Prediction from a Static Image},
  author={Jacob Walker and Abhinav Gupta and Martial Hebert},
  journal={2015 IEEE International Conference on Computer Vision (ICCV)},
  year={2015},
  pages={2443-2451}
}
  • Jacob Walker, Abhinav Gupta, Martial Hebert
  • Published 2015
  • Computer Science
  • 2015 IEEE International Conference on Computer Vision (ICCV)
  • Given a scene, what is going to move, and in what direction will it move. [...] Key Method Our CNN model leverages the data in tens of thousands of realistic videos to train our model. Our method relies on absolutely no human labeling and is able to predict motion based on the context of the scene. Because our CNN model makes no assumptions about the underlying scene, it can predict future optical flow on a diverse set of scenarios. We outperform all previous approaches by large margins.Expand Abstract

    Citations

    Publications citing this paper.
    SHOWING 1-10 OF 140 CITATIONS

    Predicting Future Optical Flow from Static Video Frames

    VIEW 9 EXCERPTS
    CITES BACKGROUND, METHODS & RESULTS
    HIGHLY INFLUENCED

    Dynamic Visual Sequence Prediction with Motion Flow Networks

    VIEW 5 EXCERPTS
    CITES METHODS & BACKGROUND
    HIGHLY INFLUENCED

    Geometry-based next frame prediction from monocular video

    VIEW 1 EXCERPT
    CITES METHODS

    Visual Dynamics: Stochastic Future Generation via Layered Cross Convolutional Networks

    VIEW 1 EXCERPT
    CITES BACKGROUND

    FILTER CITATIONS BY YEAR

    2016
    2020

    CITATION STATISTICS

    • 14 Highly Influenced Citations

    • Averaged 27 Citations per year from 2018 through 2020

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 38 REFERENCES

    Designing deep networks for surface normal estimation

    Predicting Object Dynamics in Scenes

    Large-Scale Video Classification with Convolutional Neural Networks

    VIEW 1 EXCERPT

    Patch to the Future: Unsupervised Visual Prediction

    VIEW 2 EXCERPTS

    PANDA: Pose Aligned Networks for Deep Attribute Modeling

    VIEW 1 EXCERPT

    Action Recognition with Improved Trajectories