Segmentation and Recognition Using Structure from Motion Point Clouds

@inproceedings{Brostow2008SegmentationAR,
  title={Segmentation and Recognition Using Structure from Motion Point Clouds},
  author={Gabriel J. Brostow and Jamie Shotton and Julien Fauqueur and Roberto Cipolla},
  booktitle={ECCV},
  year={2008}
}
  • Gabriel J. Brostow, Jamie Shotton, +1 author Roberto Cipolla
  • Published in ECCV 2008
  • Computer Science
  • We propose an algorithm for semantic segmentation based on 3D point clouds derived from ego-motion. We motivate five simple cues designed to model specific patterns of motion and 3D world structure that vary with object category. We introduce features that project the 3D cues back to the 2D image plane while modeling spatial layout and context. A randomized decision forest combines many such features to achieve a coherent 2D segmentation and recognize the object categories present. Our main… CONTINUE READING

    Create an AI-powered research feed to stay up to date with new papers like this posted to ArXiv

    Citations

    Publications citing this paper.
    SHOWING 1-10 OF 511 CITATIONS

    AMNet: Convolutional Neural Network embeded with Attention Mechanism for Semantic Segmentation

    VIEW 9 EXCERPTS
    CITES BACKGROUND
    HIGHLY INFLUENCED

    Attention-Guided Network for Semantic Video Segmentation

    VIEW 10 EXCERPTS
    CITES METHODS
    HIGHLY INFLUENCED

    Exploring Temporal Information for Improved Video Understanding

    VIEW 6 EXCERPTS
    CITES METHODS & BACKGROUND
    HIGHLY INFLUENCED

    Real-Time Semantic Segmentation via Multiply Spatial Fusion Network

    VIEW 10 EXCERPTS
    CITES METHODS
    HIGHLY INFLUENCED

    Learning Deep Spatio-Temporal Dependence for Semantic Video Segmentation

    VIEW 19 EXCERPTS
    CITES METHODS
    HIGHLY INFLUENCED

    Real-Time Image Semantic Segmentation Networks with Residual Depth-Wise Separable Blocks

    VIEW 4 EXCERPTS
    CITES BACKGROUND
    HIGHLY INFLUENCED

    Real-time Semantic Segmentation for Road Scene

    VIEW 8 EXCERPTS
    CITES METHODS
    HIGHLY INFLUENCED

    DAG-Recurrent Neural Networks for Scene Labeling

    VIEW 6 EXCERPTS
    CITES METHODS & RESULTS
    HIGHLY INFLUENCED

    FILTER CITATIONS BY YEAR

    2009
    2020

    CITATION STATISTICS

    • 107 Highly Influenced Citations

    • Averaged 73 Citations per year from 2017 through 2019

    • 40% Increase in citations per year in 2019 over 2018

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 30 REFERENCES

    Semantic texton forests for image categorization and segmentation

    VIEW 6 EXCERPTS

    3D LayoutCRF for Multi-View Object Class Recognition and Segmentation

    3D Model based Object Class Detection in An Arbitrary View

    3D generic object categorization, localization and pose estimation

    An Exemplar Model for Learning Object Classes

    Describing Composite Urban Workspaces

    VIEW 1 EXCERPT

    Dynamic 3D Scene Analysis from a Moving Vehicle

    VIEW 2 EXCERPTS

    Flexible Object Models for Category-Level 3D Object Recognition

    Real-Time Visibility-Based Fusion of Depth Maps

    VIEW 1 EXCERPT

    Tree-based Classifiers for Bilayer Video Segmentation

    VIEW 1 EXCERPT