Journal
IMAGE AND VISION COMPUTING
Volume 75, Issue -, Pages 21-31Publisher
ELSEVIER
DOI: 10.1016/j.imavis.2018.04.004
Keywords
Error correcting output codes; Output embeddings; Deep learning; Computer vision
Categories
Funding
- Spanish project [TIN2015-65464-R]
- Secretaria d'Universitats i Recerca del Departament d'Economia i Coneixement de la Generalitat de Catalunya [2016FI_B 01163]
- COST Action [IC1307]
- COST (European Cooperation in Science and Technology)
- NVIDIA Corporation
- GTX TITAN GPU
Ask authors/readers for more resources
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. (C) 2018 Elsevier B.V. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available