CompactNets: Compact Hierarchical Compositional Networks for Visual Recognition. Lobel, H., Vidal, R., & Soto, A. Computer Vision and Image Understanding, 191:102841, 2020.
CompactNets: Compact Hierarchical Compositional Networks for Visual Recognition [link]Paper  doi  abstract   bibtex   2 downloads  
CNN-based models currently provide state-of-the-art performance in image categorization tasks. While these methods are powerful in terms of representational capacity, they are generally not conceived with explicit means to control complexity. This might lead to scenarios where resources are used in a non-optimal manner, increasing the number of unspecialized or repeated neurons, and overfitting to data. In this work we propose CompactNets, a new approach to visual recognition that learns a hierarchy of shared, discriminative, specialized, and compact representations. CompactNets naturally capture the notion of compositional compactness, a characterization of complexity in compositional models, consisting on using the smallest number of patterns to build a suitable visual representation. We employ a structural regularizer with group-sparse terms in the objective function, that induces on each layer, an efficient and effective use of elements from the layer below. In particular, this allows groups of top-level features to be specialized based on category information. We evaluate CompactNets on the ILSVRC12 dataset, obtaining compact representations and competitive performance, using an order of magnitude less parameters than common CNN-based approaches. We show that CompactNets are able to outperform other group-sparse-based approaches, in terms of performance and compactness. Finally, transfer-learning experiments on small-scale datasets demonstrate high generalization power, providing remarkable categorization performance with respect to alternative approaches.
@Article{	  lobel2020102841,
  title		= {CompactNets: Compact Hierarchical Compositional Networks
		  for Visual Recognition},
  journal	= {Computer Vision and Image Understanding},
  volume	= {191},
  pages		= {102841},
  year		= {2020},
  issn		= {1077-3142},
  doi		= {https://doi.org/10.1016/j.cviu.2019.102841},
  url		= {https://www.sciencedirect.com/science/article/pii/S1077314218301905},
  author	= {Hans Lobel and René Vidal and Alvaro Soto},
  keywords	= {Deep learning, Regularization, Group sparsity, Image
		  categorization},
  abstract	= {CNN-based models currently provide state-of-the-art
		  performance in image categorization tasks. While these
		  methods are powerful in terms of representational capacity,
		  they are generally not conceived with explicit means to
		  control complexity. This might lead to scenarios where
		  resources are used in a non-optimal manner, increasing the
		  number of unspecialized or repeated neurons, and
		  overfitting to data. In this work we propose CompactNets, a
		  new approach to visual recognition that learns a hierarchy
		  of shared, discriminative, specialized, and compact
		  representations. CompactNets naturally capture the notion
		  of compositional compactness, a characterization of
		  complexity in compositional models, consisting on using the
		  smallest number of patterns to build a suitable visual
		  representation. We employ a structural regularizer with
		  group-sparse terms in the objective function, that induces
		  on each layer, an efficient and effective use of elements
		  from the layer below. In particular, this allows groups of
		  top-level features to be specialized based on category
		  information. We evaluate CompactNets on the ILSVRC12
		  dataset, obtaining compact representations and competitive
		  performance, using an order of magnitude less parameters
		  than common CNN-based approaches. We show that CompactNets
		  are able to outperform other group-sparse-based approaches,
		  in terms of performance and compactness. Finally,
		  transfer-learning experiments on small-scale datasets
		  demonstrate high generalization power, providing remarkable
		  categorization performance with respect to alternative
		  approaches.}
}
Downloads: 2
About
Documentation
Keywords
Search
Users
Terms and Conditions of Use
Privacy Policy
Sitemap
© BibBase.org 2021