[HTML][HTML] A survey on symmetrical neural network architectures and applications
O Ilina, V Ziyadinov, N Klenov, M Tereshonok - Symmetry, 2022 - mdpi.com
A number of modern techniques for neural network training and recognition enhancement
are based on their structures' symmetry. Such approaches demonstrate impressive results …
are based on their structures' symmetry. Such approaches demonstrate impressive results …
Siamese neural networks: An overview
D Chicco - Artificial neural networks, 2021 - Springer
Similarity has always been a key aspect in computer science and statistics. Any time two
element vectors are compared, many different similarity approaches can be used …
element vectors are compared, many different similarity approaches can be used …
A new approach to design symmetry invariant neural networks
P Kicki, P Skrzypczyński, M Ozay - 2021 International Joint …, 2021 - ieeexplore.ieee.org
We investigate a new method to design G-invariant neural networks that approximate
functions invariant to the action of a given permutation subgroup G of the symmetric group …
functions invariant to the action of a given permutation subgroup G of the symmetric group …
On the emergence of simplex symmetry in the final and penultimate layers of neural network classifiers
E Weinan, S Wojtowytsch - Mathematical and Scientific …, 2022 - proceedings.mlr.press
A recent numerical study observed that neural network classifiers enjoy a large degree of
symmetry in the penultimate layer. Namely, if $ h (x)= Af (x)+ b $ where $ A $ is a linear map …
symmetry in the penultimate layer. Namely, if $ h (x)= Af (x)+ b $ where $ A $ is a linear map …
[HTML][HTML] Optimal design of convolutional neural network architectures using teaching–learning-based optimization for image classification
Convolutional neural networks (CNNs) have exhibited significant performance gains over
conventional machine learning techniques in solving various real-life problems in …
conventional machine learning techniques in solving various real-life problems in …
On the emergence of simplex symmetry in the final and penultimate layers of neural network classifiers
S Wojtowytsch - arXiv preprint arXiv:2012.05420, 2020 - arxiv.org
A recent numerical study observed that neural network classifiers enjoy a large degree of
symmetry in the penultimate layer. Namely, if $ h (x)= Af (x)+ b $ where $ A $ is a linear map …
symmetry in the penultimate layer. Namely, if $ h (x)= Af (x)+ b $ where $ A $ is a linear map …
Learning invariant representations for equivariant neural networks using orthogonal moments
The convolutional layers of standard convolutional neural networks (CNNs) are equivariant
to translation. However, the convolution and fully-connected layers are not equivariant or …
to translation. However, the convolution and fully-connected layers are not equivariant or …
Co-attentive equivariant neural networks: Focusing equivariance on transformations co-occurring in data
DW Romero, M Hoogendoorn - arXiv preprint arXiv:1911.07849, 2019 - arxiv.org
Equivariance is a nice property to have as it produces much more parameter efficient neural
architectures and preserves the structure of the input through the feature mapping. Even …
architectures and preserves the structure of the input through the feature mapping. Even …
[HTML][HTML] Off the mainstream: advances in neural networks and machine learning for pattern recognition
This Special Issue (SI) originates from an event we organized in Ulm, Germany, in
September 2016, namely the seventh IAPR TC3 Workshop on Artificial Neural Networks in …
September 2016, namely the seventh IAPR TC3 Workshop on Artificial Neural Networks in …
[PDF][PDF] Deep symmetry networks
R Gens, PM Domingos - Advances in neural information …, 2014 - proceedings.neurips.cc
The chief difficulty in object recognition is that objects' classes are obscured by a large
number of extraneous sources of variability, such as pose and part deformation. These …
number of extraneous sources of variability, such as pose and part deformation. These …