UberNet: Training a Universal Convolutional Neural Network for Low-, Mid-, and High-Level Vision Using Diverse Datasets and Limited Memory

@article{Kokkinos2017UberNetTA,
  title={UberNet: Training a Universal Convolutional Neural Network for Low-, Mid-, and High-Level Vision Using Diverse Datasets and Limited Memory},
  author={Iasonas Kokkinos},
  journal={2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2017},
  pages={5454-5463}
}
In this work we train in an end-to-end manner a convolutional neural network (CNN) that jointly handles low-, mid-, and high-level vision tasks in a unified architecture. Such a network can act like a swiss knife for vision tasks, we call it an UberNet to indicate its overarching nature. The main contribution of this work consists in handling challenges that emerge when scaling up to many tasks. We introduce techniques that facilitate (i) training a deep architecture while relying on diverse… CONTINUE READING
Highly Cited
This paper has 129 citations. REVIEW CITATIONS
Recent Discussions
This paper has been referenced on Twitter 38 times over the past 90 days. VIEW TWEETS

Citations

Publications citing this paper.
Showing 1-10 of 92 extracted citations

130 Citations

050100201620172018
Citations per Year
Semantic Scholar estimates that this publication has 130 citations based on the available data.

See our FAQ for additional information.

References

Publications referenced by this paper.
Showing 1-10 of 106 references

Similar Papers

Loading similar papers…