Corpus ID: 5344420

Re-ranking Object Proposals for Object Detection in Automatic Driving

@article{Zhong2016RerankingOP,
  title={Re-ranking Object Proposals for Object Detection in Automatic Driving},
  author={Z. Zhong and M. Lei and Shaozi Li and Jianping Fan},
  journal={ArXiv},
  year={2016},
  volume={abs/1605.05904}
}
  • Z. Zhong, M. Lei, +1 author Jianping Fan
  • Published 2016
  • Computer Science
  • ArXiv
  • Object detection often suffers from a plenty of bootless proposals, selecting high quality proposals remains a great challenge. In this paper, we propose a semantic, class-specific approach to re-rank object proposals, which can consistently improve the recall performance even with less proposals. We first extract features for each proposal including semantic segmentation, stereo information, contextual information, CNN-based objectness and low-level cue, and then score them using class… CONTINUE READING
    Unsupervised action proposal ranking through proposal recombination
    3
    Feature++: Cross dimension feature fusion for road detection
    1
    Artificial intelligence applications in the development of autonomous vehicles: a survey
    3

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 28 REFERENCES
    What Makes for Effective Detection Proposals?
    521
    Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
    11029
    DeepBox: Learning Objectness with Convolutional Networks
    138
    Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
    17487
    3D Object Proposals for Accurate Object Class Detection
    390
    Measuring the Objectness of Image Windows
    1093
    BING: Binarized Normed Gradients for Objectness Estimation at 300fps
    605
    Scalable Object Detection Using Deep Neural Networks
    750
    Convolutional feature masking for joint object and stuff segmentation
    323
    Segmentation as selective search for object recognition
    607