@Article{PauRodriguez2018,
author="Pau Rodriguez
and Miguel Angel Bautista
and Sergio Escalera
and Jordi Gonzalez",
title="Beyond Oneshot Encoding: lower dimensional target embedding",
journal="Image and Vision Computing",
year="2018",
volume="75",
pages="21--31",
optkeywords="Error correcting output codes",
optkeywords="Output embeddings",
optkeywords="Deep learning",
optkeywords="Computer vision",
abstract="Target encoding plays a central role when learning Convolutional Neural Networks. In this realm, one-hot encoding is the most prevalent strategy due to its simplicity. However, this so widespread encoding schema assumes a flat label space, thus ignoring rich relationships existing among labels that can be exploited during training. In large-scale datasets, data does not span the full label space, but instead lies in a low-dimensional output manifold. Following this observation, we embed the targets into a low-dimensional space, drastically improving convergence speed while preserving accuracy. Our contribution is two fold: (i) We show that random projections of the label space are a valid tool to find such lower dimensional embeddings, boosting dramatically convergence rates at zero computational cost; and (ii) we propose a normalized eigenrepresentation of the class manifold that encodes the targets with minimal information loss, improving the accuracy of random projections encoding while enjoying the same convergence rates. Experiments on CIFAR-100, CUB200-2011, Imagenet, and MIT Places demonstrate that the proposed approach drastically improves convergence speed while reaching very competitive accuracy rates.",
optnote="ISE; HuPBA; 600.098; 602.133; 602.121; 600.119",
optnote="exported from refbase (http://refbase.cvc.uab.es/show.php?record=3120), last updated on Thu, 16 Feb 2023 12:03:17 +0100",
doi="10.1016/j.imavis.2018.04.004",
opturl="https://doi.org/10.1016/j.imavis.2018.04.004",
file=":http://refbase.cvc.uab.es/files/RBE2018.pdf:PDF"
}