Currently, video generation devices are simpler to manipulate, more portable and with lower prices. This allowed easy storage and transmission of large amounts of media, such as videos, which has facilitated the analysis of information, independent of human assistance for evaluation and exhaustive search of videos. Virtual reality, robotics, tele-medicine, humanmachine interface and tele-surveillance are applications for these techniques. This paper describes a method for human action recognition in videos using two convolutional neural networks (CNNs). The first one Spatial Stream (trained with frames of the video) and the second one Temporal Stream, trained with stacks of Dense Optical Flow (DOF). Both streams were trained separately and from both of them we generated a classification histogram based on the most frequent class assignment. For final classification, those histograms were combined to produce a single output. The technique was tested in two public action video datasets: Weizmann and UCF Sports. We achieve 84.44% of accuracy on Weizmann dataset for Spatial stream and 78.46% on UCF Sports dataset. For the Weizmann dataset we obtained 91.11% with networks combination.

This paper describes the development and application of a method for human action recognition from motion analysis in a sequence of images using an artificial neural network. The proposed method is based on two stages: Computer Vision and Computational Intelligence. The Computer Vision stage is a combination of two motion analysis techniques: Histogram of Oriented Optical Flow and Object Contour Analysis. For the Computational Intelligence stage we use a Self-Organizing Map (SOM) optimized through Learning Vector Quantization (LVQ). The approach is then applied for classification of human actions in many real situations. Testing against a database with different kinds of human actions, we show the usefulness and robustness of this method, comparing it to other proposals in the literature.

This work aims to present a device for the objective evaluation of video quality from the user perspective, in mass content distribution networks. This simple and cost-effective device with an integrated image sensor and an embedded processor intends to be network pervasive while minimizing the network overhead, allowing quality measurements over the network with greater control of the content distribution throughout the video chain, from content producers, distributors and to consumers.

In this paper, the kinematic model of a quadruped robot is derived. The model is equivalent to that of a parallel manipulator, in that each leg can be seen as a manipulator. However, the model is extended to consider that in one gait cycle some legs are in contact with the ground and others are not. In order to obtain the inverse kinematics model, this paper presents as contribution the use of the extended Kalman filter as optimizer in two different situations of the leg motion: unconstrained case, for the swing leg(s), and constrained case, for the leg(s) in contact with the ground. This method was evaluated for locomotion in plain and inclined surfaces. The results obtained with the kinematics model were satisfactory when implemented in a point-to-point trajectory in simulation, and also in an experiment with a four-legged platform with three degrees of freedom in each leg.

This paper presents a statistical analysis of stochastic-based optimization algorithms applied to a digital filter arbitrary magnitude and phase approximation design problem. Using an already developed rigorous statistical methodology, a completely randomized design is set up and best parameters values are estimated for the adaptive algorithms applied to a specific non-linear approximation problem. After finding the best parameter values, an additional completely randomized design is set up, comparing the performance of the adaptive algorithms with a quasi-Newton algorithm. Results for the statistical analysis are presented and the performance for different optimization algorithms with the best parameter values are analyzed.

In this paper we use computational intelligence numerical optimization techniques to design optimum non-linear phase finite impulse response digital filters. The filter magnitude and group delay approximation will be presented as a classical approximation problem, where a non-linear function of several variables must be minimized. Results will be presented and compared with three different numerical optimization techniques developed, including Genetic algorithms, Simulated Annealing and Particle Swarm Optimization. Additionally, results are compared with a current used design technique, an unconstrained quasi-Newton algorithm.

An optoelectronic implementation of a spiking neuron model based on the FitzHugh-Nagumo equations is presented. A tunable semiconduc- tor laser source and a spectral filter provide a nonlinear mapping from driver voltage to detected signal. Linear electronic feedback completes the implementation, which allows either electronic or optical input sig- nals. Experimental results for a single system and numeric results of model interaction confirm that important features of spiking neural mod- els can be implemented through this approach.