Learning to Label Aerial Images from Noisy Data

@inproceedings{Mnih2012LearningTL,
  title={Learning to Label Aerial Images from Noisy Data},
  author={Volodymyr Mnih and Geoffrey E. Hinton},
  booktitle={ICML},
  year={2012}
}
When training a system to label images, the amount of labeled training data tends to be a limiting factor. We consider the task of learning to label aerial images from existing maps. These provide abundant labels, but the labels are often incomplete and sometimes poorly registered. We propose two robust loss functions for dealing with these kinds of label noise and use the loss functions to train a deep neural network on two challenging aerial image datasets. The robust loss functions lead to… CONTINUE READING
Highly Influential
This paper has highly influenced 11 other papers. REVIEW HIGHLY INFLUENTIAL CITATIONS
Highly Cited
This paper has 198 citations. REVIEW CITATIONS

From This Paper

Figures, tables, and topics from this paper.

Citations

Publications citing this paper.
Showing 1-10 of 130 extracted citations

Stacked U-Nets With Multi-Output for Road Extraction

CVPR Workshops • 2018
View 10 Excerpts
Highly Influenced

Training deep neural-networks based on unreliable labels

2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) • 2016
View 5 Excerpts
Highly Influenced

Deep Learning from Noisy Image Labels with Quality Embedding

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society • 2018
View 6 Excerpts
Highly Influenced

Semiautomatic Road Extraction From VHR Images Based on Multiscale and Spectral Angle in Case of Earthquake

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing • 2018
View 3 Excerpts
Highly Influenced

An automatic pipeline for mapping roads from aerial images

AutonomousGIS@SIGSPATIAL • 2017
View 4 Excerpts
Highly Influenced

Making Deep Neural Networks Robust to Label Noise: A Loss Correction Approach

2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) • 2017
View 4 Excerpts
Highly Influenced

198 Citations

0204060'14'16'18
Citations per Year
Semantic Scholar estimates that this publication has 198 citations based on the available data.

See our FAQ for additional information.

References

Publications referenced by this paper.
Showing 1-10 of 16 references

Similar Papers

Loading similar papers…