Bruno J. T. Fernandes

Lateral Inhibition Pyramidal Neural Network for Image Classification

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 receptive and inhibitory fields. The first model is a pyramidal neural network with lateral inhibition, called lateral inhibition pyramidal neural network. The second proposed model is a supervised image segmentation system, called segmentation and classification based on receptive fields. This work shows that the combination of these two models is beneficial, and the results obtained are better than that of other state-of-the-art methods.

Classification and Segmentation of Visual Patterns Based on Receptive and Inhibitory Fields

This paper presents a new model to realize a supervised image segmentation task. It is based on the concept of receptive fields that intends to analyze pieces of an image considering not only the pixels or group of them, but also the relationship between them and their neighbors, called segmentation and classification with receptive fields (SCRF). Also, in order to work with the SCRF model, is proposed here a new artificial neural network, called IPyraNet, which is a hybrid implementation of the recently described PyraNet and the nonclassical receptive fields inhibition. Furthermore, the model and the network are applied together in order to realize a satellite image segmentation task.

HAGR-D: A Novel Approach for Gesture Recognition with Depth Maps

The hand is an important part of the body used to express information through gestures, and its movements can be used in dynamic gesture recognition systems based on computer vision with practical applications, such as medical, games and sign language. Although depth sensors have led to great progress in gesture recognition, hand gesture recognition still is an open problem because of its complexity, which is due to the large number of small articulations in a hand. This paper proposes a novel approach for hand gesture recognition with depth maps generated by the Microsoft Kinect Sensor (Microsoft, Redmond, WA, USA) using a variation of the CIPBR (convex invariant position based on RANSAC) algorithm and a hybrid classifier composed of dynamic time warping (DTW) and Hidden Markov models (HMM), called the hybrid approach for gesture recognition with depth maps (HAGR-D). The experiments show that the proposed model overcomes other algorithms presented in the literature in hand gesture recognition tasks, achieving a classification rate of 97.49% in the MSRGesture3D dataset and 98.43% in the RPPDI dynamic gesture dataset.

Convexity local contour sequences for gesture recognition

Algorithms for hand feature extraction used in gesture recognition systems have some problems such as unnecessary information gathering. This paper proposes a novel method for feature extraction in gesture recognition systems based on the Local Contour Sequence (LCS). It is called the Convexity Local Contour Sequence (CLCS) and represents the hand shape only with the most significant information. This generates a smaller output result, but capable to model an entire dynamic gesture. It is used to classify dynamic gestures with an Elman Recurrent Network and Hidden Markov Model and presents a better result compared to regular LCS.

An Effective Dynamic Gesture Recognition System Based on the Feature Vector Reduction for SURF and LCS

Speed Up Robust Feature (SURF) and Local Contour Sequence(LCS) are methods used for feature extraction techniques for dynamic gesture recognition. A problem presented by these techniques is the large amount of data in the output vector which difficult the classification task. This paper presents a novel method for dimensionality reduction of the features extracted by SURF and LCS, called Convexity Approach. The proposed method is evaluated in a gesture recognition task and improves the recognition rate of LCS while SURF while decreases the amount of data in the output vector.

AutoAssociative Pyramidal Neural Network for one class pattern classification with implicit feature extraction

Receptive fields and autoassociative memory are brain concepts that have individually inspired many artificial models, but models using both ideas have not been deeply studied. In this paper, we propose the AutoAssociative Pyramidal Neural Network (AAPNet), which is an artificial neural network for one-class classification that uses autoassociative memory and receptive field concepts in its pyramidal architecture. The proposed neural network performs implicit feature extraction and learns how to reconstruct a pattern from such features. The AAPNet is evaluated using the object categorization Caltech-101 database and presents better results when compared with other state-of-the-art methods.

Understanding how deep neural networks learn face expressions

Deep neural networks have been used successfully for several different computer vision-related tasks, including facial expression recognition. In spite of the good results, it is still not clear why these networks achieve such good recognition rates. One way to learn more about deep neural networks is to visualise and understand what they are learning, and to do so techniques such as deconvolution could play a significant role. In this paper, we train a Convolutional Neural Network (CNN) and Lateral Inhibition Pyramidal Neural Network (LIPNet) to learn facial expressions. Then, we use the deconvolution process to visualise the learned features of the CNN and we introduce a novel mechanism for visualising the internal representation of the LIPNet. We perform a series of experiments, training our networks with the Cohn-Kanade data set and show what kind of facial structures compose the learned emotion expression representation. Then, we use the trained networks to recognise images from the Jaffe data set and demonstrate that the learned representations are present in different face images, emphasizing the generalization aspects of these networks. We discuss the different representations that each network learns and how they differ from each other. We also discuss how each learned representation contributes to the recognition process and how they can be compared to the emotional notation Facial Action Coding System - Facs. Finally, we explain how the principles of invariance, redundancy and filtering, common for deep networks, contribute to the learned features and to the facial expression recognition task in general.

Lateral Inhibition Pyramidal Neural Networks Designed by Particle Swarm Optimization

LIPNet is a pyramidal neural network with lateral inhibition developed for pattern recognition, inspired in the concept of receptive and inhibitory fields from the human visual system. Although this network can implicitly extract features and use these features to properly classify patterns in images, many parameters must be defined prior to the network training and operation. Besides, these parameters have a huge impact on the recognition performance. This paper proposes an encoding scheme aiming at optimizing the LIPNet structure using Particle Swarm Optimization. Preliminary results for a face detection problem using a well known benchmark set showed that our approach achieved better classification rates when compared to the original LIPNet.

A Dynamic Gesture Recognition System based on CIPBR Algorithm

Dynamic gesture recognition has been studied actually for it big application in several areas such as virtual reality, games and sign language. But some problems have to be solved in computer applications, such as response time and classification rate, which directly affect the real-time usage. This paper proposes a novel algorithm called Convex Invariant Position Based on Ransac which improved the good results in dynamic gesture recognition problem. The proposed method is combined with a adapted PSO variation to reduce features and a HMM and three DTW variations as classifiers. Index Terms—Gesture recognition, computer vision, CIPBR, dynamic time wrapping, hidden Markov model.

An AutoAssociative Neural Network for image segmentation

In this paper, it is proposed a neural network based on by AutoAssociative Pyramidal Neural Network and their architecture, which uses concepts of receptive fields and autoassociative memory. These concepts are widely used in models of artificial neural networks and were incorporated into model proposed in this work. Furthermore, the proposed neural network also uses the concept of sharing weights aiming the applications on problems invariant translations. The neural network is able of perform implicit feature extraction and learns how to reconstruct a pattern of such features. The evaluation of the neural network is performed by two experiments. The first experiment is conducted with image processing problems. The Neural Network Autoassociative learns about the transformation applied to the images, mapping a domain of images to another. In the second experiment the AutoAssociative Neural Network gets satisfactory results in image segmentation. The second task uses the dataset of skin lesion images for segmentation. This work indicates that proposed model of neural network is valid due to the obtained results achieved in the performed experiments.