Lateral Inhibition Pyramidal Neural Network for Image Classification

@article{Fernandes2013LateralIP,
  title={Lateral Inhibition Pyramidal Neural Network for Image Classification},
  author={Bruno Jos{\'e} Torres Fernandes and George D. C. Cavalcanti and Ing Ren Tsang},
  journal={IEEE Transactions on Cybernetics},
  year={2013},
  volume={43},
  pages={2082-2092}
}
The human visual system is one of the most fascinating and complex mechanisms of the central nervous system that enables our capacity to see. It is through the visual system that we are able to accomplish from the most simple task such as object recognition to the most complex visual interpretation, understanding and perception. Inspired by this sophisticated system, two models based on the properties of the human visual system are proposed. These models are designed based on the concepts of… 
Non-negative pyramidal neural network for parts-based learning
TLDR
The results show that the LIPNet with non-negative constraint presented a better interpretability of the patterns, and a display model of the network learning is proposed, which is possible to compare the internal representation of the non- negative model learning and the traditional LIP net model.
Pyramidal neural networks with evolved variable receptive fields
TLDR
A PNN with variable receptive fields determined by an evolutionary algorithm, called variable pyramidal neural network with evolutionary algorithms is proposed, which can achieve better classification rates and is comparable to biological systems regarding the flexibility assigned to receptive fields.
Lateral Inhibition Pyramidal Neural Networks Designed by Particle Swarm Optimization
TLDR
This paper proposes an encoding scheme aiming at optimizing the LIPNet structure using Particle Swarm Optimization, and preliminary results for a face detection problem showed that this approach achieved better classification rates when compared to the original LIP net.
Structured Pyramidal Neural Networks
TLDR
A variation of PNN, named here as Structured Pyramidal Neural Network (SPNN), which has self-adaptive variable receptive fields and has better results than PNNs and similar performance when compared to Convolutional Neural Network and Support Vector Machine, but using lower memory capacity and processing time.
Pyramidal Neural Networks with Variable Receptive Fields Designed by Genetic Algorithms
TLDR
This paper proposes a PNN with variable receptive fields, which might be determined by a Genetic Algorithm, called Variable Pyramidal Neural Network with Genetic Algorithms (VPNN-GA).
Lateral Inhibition-Inspired Convolutional Neural Network for Visual Attention and Saliency Detection
TLDR
This paper proposes to formulate lateral inhibition inspired by the related studies from neurobiology, and embed it into the top-down gradient computation of a general CNN for classification, i.e. only category-level information is used.
Non-negative Structured Pyramidal Neural Network for Pattern Recognition
TLDR
A non-negative model of the SPNN is proposed, to obtain better interpretability and stability of the network learning, and is evaluated in a gender recognition problem using the Face Recognition Technology (FERET) database.
Enhancing Semantic Segmentation of Aerial Images with Inhibitory Neurons
TLDR
This paper begins with deep network architectures such as VGG and ResNet, and proposes an approach to add lateral inhibition in each output neuron to reduce its impact on its neighbours, both in fine-tuning pre-trained models and training from scratch.
Constructive Autoassociative Neural Network for Facial Recognition
TLDR
Experiments in face recognition and facial expression recognition show that the CANet outperforms other methods presented in the literature.
Biologically Inspired Semantic Lateral Connectivity for Convolutional Neural Networks
TLDR
It is shown that establishing a biologically inspired Mexican hat lateral connectivity profile along the filter domain can significantly improve the classification accuracy of a variety of lightweight convolutional neural networks without the addition of trainable network parameters.
...
...

References

SHOWING 1-10 OF 47 REFERENCES
Classification and Segmentation of Visual Patterns Based on Receptive and Inhibitory Fields
TLDR
A new artificial neural network is proposed here, called IPyraNet, which is a hybrid implementation of the recently described PyraNet and the nonclassical receptive fields inhibition, and applied together in order to realize a satellite image segmentation task.
Spatio–Temporal Adaptation in the Unsupervised Development of Networked Visual Neurons
TLDR
This paper presents a new model called neuronal cluster which includes spatial as well as temporal weights in its unified adaptation scheme, and presents the mathematical demonstration that the temporal weights are derived from the delay in lateral inhibition.
A neurocomputational model of figure-ground discrimination and target tracking
TLDR
It is shown that through the oscillatory amplitude and frequency encoding, and selective synchronization of phase oscillators, the figure and the ground can be successfully discriminated from each other.
A Pyramidal Neural Network For Visual Pattern Recognition
TLDR
This paper applies PyraNet to determine gender from a facial image, and compares its performance on the standard facial recognition technology (FERET) database with three classifiers: The convolutional neural network (NN), the k-nearest neighbor (k-NN), and the support vector machine (SVM).
Contour detection based on nonclassical receptive field inhibition
TLDR
The proposed biologically motivated method to improve contour detection in machine vision, called nonclassical receptive field (non-CRF) inhibition (more generally, surround inhibition or suppression), is proposed and is more useful for contour-based object recognition tasks, than traditional edge detectors, which do not distinguish between contour and texture edges.
Neocognitron: A neural network model for a mechanism of visual pattern recognition
TLDR
A recognition with a large-scale network is simulated on a PDP-11/34 minicomputer and is shown to have a great capability for visual pattern recognition and can be trained to recognize handwritten Arabic numerals even with considerable deformations in shape.
Neocognitron: A hierarchical neural network capable of visual pattern recognition
Nonclassical Receptive Field Inhibition Applied to Image Segmentation
TLDR
A new artiflcial neural network, called I-PyraNet, is proposed, which is a hybrid implementation of the recently described PyraNet and the nonclassical receptive flelds inhibition to accomplish a satellite image segmentation task.
Neural Networks for Pattern Recognition
Efficient training algorithms for a class of shunting inhibitory convolutional neural networks
TLDR
The modified Broyden-Fletcher-Goldfarb-Shanno (BFGS) methods, the three variants of LM algorithm, and the new hybrid/LS method perform consistently well, achieving error rates of less than 3% averaged across all three architectures.
...
...