Advanced search
1 file | 1.80 MB Add to list
Author
Organization
Abstract
We present a neural process that models exchangeable sequences of high-dimensional complex observations conditionally on a set of labels or tags. Our model combines the expressiveness of deep neural networks with the data-efficiency of Gaussian processes, resulting in a probabilistic model for which the posterior distribution is easy to evaluate and sample from, and the computational complexity scales linearly with the number of observations. The advantages of the proposed architecture are demonstrated on a challenging few-shot view reconstruction task which requires generalisation from short sequences of viewpoints.

Downloads

  • es2019-35.pdf
    • full text (Published version)
    • |
    • open access
    • |
    • PDF
    • |
    • 1.80 MB

Citation

Please use this url to cite or link to this publication:

MLA
Korshunova, Iryna, et al. “Conditional BRUNO : A Neural Process for Exchangeable Labelled Data.” Proceedings of the 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2019), ESANN, 2019, pp. 1–6.
APA
Korshunova, I., Gal, Y., Gretton, A., & Dambre, J. (2019). Conditional BRUNO : a neural process for exchangeable labelled data. In Proceedings of the 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2019) (pp. 1–6). Bruges, Belgium: ESANN.
Chicago author-date
Korshunova, Iryna, Yarin Gal, Arthur Gretton, and Joni Dambre. 2019. “Conditional BRUNO : A Neural Process for Exchangeable Labelled Data.” In Proceedings of the 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2019), 1–6. ESANN.
Chicago author-date (all authors)
Korshunova, Iryna, Yarin Gal, Arthur Gretton, and Joni Dambre. 2019. “Conditional BRUNO : A Neural Process for Exchangeable Labelled Data.” In Proceedings of the 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2019), 1–6. ESANN.
Vancouver
1.
Korshunova I, Gal Y, Gretton A, Dambre J. Conditional BRUNO : a neural process for exchangeable labelled data. In: Proceedings of the 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2019). ESANN; 2019. p. 1–6.
IEEE
[1]
I. Korshunova, Y. Gal, A. Gretton, and J. Dambre, “Conditional BRUNO : a neural process for exchangeable labelled data,” in Proceedings of the 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2019), Bruges, Belgium, 2019, pp. 1–6.
@inproceedings{8613329,
  abstract     = {We present a neural process that models exchangeable sequences of high-dimensional complex observations conditionally on a set of labels or tags. Our model combines the expressiveness of deep neural networks with the data-efficiency of Gaussian processes, resulting in a probabilistic model for which the posterior distribution is easy to evaluate and sample from, and the computational complexity scales linearly with the number of observations. The advantages of the proposed architecture are demonstrated on a challenging few-shot view reconstruction task which requires generalisation from short sequences of viewpoints.},
  author       = {Korshunova, Iryna and Gal, Yarin and Gretton, Arthur and Dambre, Joni},
  booktitle    = {Proceedings of the 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2019)},
  isbn         = {9782875870650},
  language     = {eng},
  location     = {Bruges, Belgium},
  pages        = {1--6},
  publisher    = {ESANN},
  title        = {Conditional BRUNO : a neural process for exchangeable labelled data},
  url          = {https://www.elen.ucl.ac.be/Proceedings/esann/esannpdf/es2019-35.pdf},
  year         = {2019},
}