Corpus ID: 211296865

Circle Loss: A Unified Perspective of Pair Similarity Optimization

@article{Sun2020CircleLA,
  title={Circle Loss: A Unified Perspective of Pair Similarity Optimization},
  author={Yifan Sun and Changmao Cheng and Yuhan Zhang and Chi Zhang and Liang Zheng and Zhongdao Wang and Yichen Wei},
  journal={ArXiv},
  year={2020},
  volume={abs/2002.10857}
}
  • Yifan Sun, Changmao Cheng, +4 authors Yichen Wei
  • Published 2020
  • Computer Science
  • ArXiv
  • This paper provides a pair similarity optimization viewpoint on deep feature learning, aiming to maximize the within-class similarity $s_p$ and minimize the between-class similarity $s_n$. We find a majority of loss functions, including the triplet loss and the softmax plus cross-entropy loss, embed $s_n$ and $s_p$ into similarity pairs and seek to reduce $(s_n-s_p)$. Such an optimization manner is inflexible, because the penalty strength on every single similarity score is restricted to be… CONTINUE READING

    Citations

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

    VOC-ReID: Vehicle Re-identification based on Vehicle-Orientation-Camera

    VIEW 4 EXCERPTS
    CITES METHODS & BACKGROUND
    HIGHLY INFLUENCED

    FastReID: A Pytorch Toolbox for Real-world Person Re-identification

    VIEW 1 EXCERPT
    CITES BACKGROUND

    Generalizing Face Representation with Unlabeled Data

    VIEW 1 EXCERPT
    CITES BACKGROUND

    MISIM: An End-to-End Neural Code Similarity System

    VIEW 2 EXCERPTS
    CITES METHODS & BACKGROUND

    Quasi-Dense Instance Similarity Learning

    VIEW 1 EXCERPT
    CITES METHODS

    Ranked List Loss for Deep Metric Learning

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 42 REFERENCES

    CosFace: Large Margin Cosine Loss for Deep Face Recognition

    • Hai Jun Wang, Yitong Wang, +5 authors Wenyu Liu
    • Computer Science
    • 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
    • 2018
    VIEW 15 EXCERPTS
    HIGHLY INFLUENTIAL

    3D Object Representations for Fine-Grained Categorization

    VIEW 6 EXCERPTS
    HIGHLY INFLUENTIAL

    Deep Metric Learning via Lifted Structured Feature Embedding

    VIEW 9 EXCERPTS
    HIGHLY INFLUENTIAL

    Multi-Similarity Loss With General Pair Weighting for Deep Metric Learning

    VIEW 7 EXCERPTS
    HIGHLY INFLUENTIAL

    Additive Margin Softmax for Face Verification

    VIEW 7 EXCERPTS
    HIGHLY INFLUENTIAL

    ArcFace: Additive Angular Margin Loss for Deep Face Recognition

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    Deep Metric Learning via Facility Location

    VIEW 3 EXCERPTS
    HIGHLY INFLUENTIAL