Knowledge Adaptation for Efficient Semantic Segmentation

@article{He2019KnowledgeAF,
  title={Knowledge Adaptation for Efficient Semantic Segmentation},
  author={Tong He and Chunhua Shen and Zhi Tian and Dong Gong and Changming Sun and Youliang Yan},
  journal={2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2019},
  pages={578-587}
}
  • Tong He, Chunhua Shen, +3 authors Youliang Yan
  • Published in
    IEEE/CVF Conference on…
    2019
  • Computer Science
  • Both accuracy and efficiency are of significant importance to the task of semantic segmentation. Existing deep FCNs suffer from heavy computations due to a series of high-resolution feature maps for preserving the detailed knowledge in dense estimation. Although reducing the feature map resolution (i.e., applying a large overall stride) via subsampling operations (e.g., polling and convolution striding) can instantly increase the efficiency, it dramatically decreases the estimation accuracy. To… CONTINUE READING

    Citations

    Publications citing this paper.
    SHOWING 1-10 OF 14 CITATIONS

    Role-Wise Data Augmentation for Knowledge Distillation

    VIEW 1 EXCERPT
    CITES BACKGROUND

    Temporally Distributed Networks for Fast Video Semantic Segmentation

    VIEW 1 EXCERPT
    CITES BACKGROUND

    Decoders Matter for Semantic Segmentation: Data-Dependent Decoding Enables Flexible Feature Aggregation

    VIEW 1 EXCERPT

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 32 REFERENCES

    FitNets: Hints for Thin Deep Nets

    VIEW 8 EXCERPTS
    HIGHLY INFLUENTIAL

    Deep Residual Learning for Image Recognition

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    MobileNetV2: Inverted Residuals and Linear Bottlenecks

    VIEW 8 EXCERPTS
    HIGHLY INFLUENTIAL

    Xception: Deep Learning with Depthwise Separable Convolutions

    • François Chollet
    • Computer Science
    • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
    • 2017
    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    The Cityscapes Dataset for Semantic Urban Scene Understanding

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    The Pascal Visual Object Classes Challenge: A Retrospective

    VIEW 7 EXCERPTS
    HIGHLY INFLUENTIAL

    The Role of Context for Object Detection and Semantic Segmentation in the Wild

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs

    VIEW 6 EXCERPTS
    HIGHLY INFLUENTIAL