Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Explore this paper's citation graph

Summary

Applied to a state-of-the-art image classification model, Batch Normalization achieves the same accuracy with 14 times fewer training steps, and beats the original model by a significant margin.

Type
preprint
Published
2015-02-10
Cited by
47,615
References
54
Access
Open access

Keywords

Covariate, Normalization (sociology), Computer science, Training (meteorology), Econometrics

References

Cited by

Related papers