Search

In the last decade, exponential data growth supplied machine learning-based algorithms' capacity and enabled their usage in daily-life activities. Additionally, such an improvement is partially explained due to the advent of deep learning techniques, i.e., stacks of simple architectures that end up in more complex models. Although both factors produce outstanding results, they also pose drawbacks regarding the learning process as training complex models over large datasets are expensive and time-consuming. Such a problem is even more evident when dealing with video analysis. Some works have considered transfer learning or domain adaptation, i.e., approaches that map the knowledge from one domain to another, to ease the training burden, yet most of them operate over individual or small blocks of frames. This paper proposes a novel approach to map the knowledge from action recognition to event recognition using an energy-based model, denoted as Spectral Deep Belief Network. Such a model can process all frames simultaneously, carrying spatial and temporal information through the learning process. The experimental results conducted over two public video dataset, the HMDB-51 and the UCF-101, depict the effectiveness of the proposed model and its reduced computational burden when compared to traditional energy-based models, such as Restricted Boltzmann Machines and Deep Belief Networks.

Deep learning architectures have been widely fostered throughout the last years, being used in a wide range of applications, such as object recognition, image reconstruction, and signal processing. Nevertheless, such models suffer from a common problem known as overfitting, which limits the network from predicting unseen data effectively. Regularization approaches arise in an attempt to address such a shortcoming. Among them, one can refer to the well-known Dropout, which tackles the problem by randomly shutting down a set of neurons and their connections according to a certain probability. Therefore, this approach does not consider any additional knowledge to decide which units should be disconnected. In this paper, we propose an energy-based Dropout (E-Dropout) that makes conscious decisions whether a neuron should be dropped or not. Specifically, we design this regularization method by correlating neurons and the model's energy as an importance level for further applying it to energy-based models, such as Restricted Boltzmann Machines (RBMs). The experimental results over several benchmark datasets revealed the proposed approach's suitability compared to the traditional Dropout and the standard RBMs.

Manganites have been attracted considerable attention due to some intriguing magnetic properties, such as magnetoresistance, spin glass behavior and superparamagnetism. In recent years, some studies point to the effect of particle size and dimensionality of these compounds in their magnetic features. Particularly, LaCaMnO material research is well explored concerning the bulk material. To overcome the lack of the information we successfully produced advanced nanostructures of La$_{0.6}$Ca$_{0.4}$MnO$_{3}$ manganites, namely nanotubes and nanoparticles by using a sol-gel modified method, to determine the size particle effect on the magnetism. The manganites crystal structure, magnetic and magnetocaloric properties were studied in a broad temperature range. Transmission electron microscopy revealed nanoparticles with sizes from 45 up to 223 nm, depending on the calcination temperature. It was found that the magnetic and magnetocaloric properties can be optimized by tuning the particle size; for instance, the magnetic transition broadening by decreasing the particle size. We report the relative cooling power (RCP) of these samples; it was found that the best RCP was observed for the 223 nm particle (508 J/Kg). Finally, this work contributes to the research on the magnetic properties and magnetocaloric potentials in nanostructured systems with distinct morphologies., Comment: 11 pages, 8 figures