Corpus ID: 212747706

Fixing the train-test resolution discrepancy: FixEfficientNet

@article{Touvron2020FixingTT,
  title={Fixing the train-test resolution discrepancy: FixEfficientNet},
  author={Hugo Touvron and Andrea Vedaldi and Matthijs Douze and Herv{\'e} J{\'e}gou},
  journal={ArXiv},
  year={2020},
  volume={abs/2003.08237}
}
  • Hugo Touvron, Andrea Vedaldi, +1 author Hervé Jégou
  • Published 2020
  • Computer Science
  • ArXiv
  • This note complements the paper "Fixing the train-test resolution discrepancy" that introduced the FixRes method. First, we show that this strategy is advantageously combined with recent training recipes from the literature. Most importantly, we provide new results for the EfficientNet architecture. The resulting network, called FixEfficientNet, significantly outperforms the initial architecture with the same number of parameters. For instance, our FixEfficientNet-B0 trained without additional… CONTINUE READING

    Figures, Tables, and Topics from this paper.

    Explore Further: Topics Discussed in This Paper

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 17 REFERENCES

    Adversarial Examples Improve Image Recognition

    VIEW 2 EXCERPTS

    Randaugment: Practical automated data augmentation with a reduced search space

    VIEW 1 EXCERPT

    Learning Transferable Architectures for Scalable Image Recognition

    VIEW 1 EXCERPT

    Do ImageNet Classifiers Generalize to ImageNet?

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks

    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    Deep Residual Learning for Image Recognition

    VIEW 1 EXCERPT