São Paulo Federal Institute of Education, Science and Technology

About: São Paulo Federal Institute of Education, Science and Technology is a education organization based out in São Paulo, Brazil. It is known for research contribution in the topics: Context (language use) & Computer science. The organization has 1707 authors who have published 2374 publications receiving 11333 citations.

Performance analysis of κ-μ distribution for Global Positioning System (GPS) L1 frequency-related ionospheric fading channels

Abstract: The present work aims to evaluate the application of the κ -μ distribution as a representation of the fading effect caused by the phenomenon of scintillation on L -band transionospheric radio links. The ionospheric scintillation is a phenomenon defined as a rapid variation in the amplitude and phase of electromagnetic wave signals. This phenomenon starts in the first hours of the night, at latitudes near the geomagnetic equator. Scintillation occurs when radio signals cross ionospheric irregularities, also known as plasma bubbles. These plasma bubble structures are generated after the sunset due to instabilities in the F region of ionosphere. Distributions with non-single parameter usually present better results, however, this point requires further investigation by comparing different models. The goals of this study are: (1) the modeling of experimental data using the κ -μ distribution; (2) the κ -μ parameters characterization for empirical data and the evaluation of parameters estimation based in different approaches; (3) the comparison between the distribution proposed and other models adopted in the literature in order to verify the performance of two parameter models. The results of the analysis performed showed that the κ -μ distribution presents good fitting of the empirical scintillation data. These fitting results were calculated through the chi-square fit test under which the values reveal fair E [χ 2 ] for κ -μ distribution in most of the cases. The evaluation of κ -μ parameters suggests that the distribution has a more conservative outcome than in the distributions traditionally used, but being a legitimate approximation due to its adjustable features in the tail region of the distribution. Typical pairs of κ -μ coefficients are presented for theoretical works. The comparison of κ -μ distribution to Rice, Nakagami-m and α -μ models showed that κ-μ is capable of describing more severe scintillation scenarios where the tail of the distribution is more raised in comparison to the other models.

Oxygen Transfer and Fragmentation of Aspergillus niger Pellets in Stirred Tank and Concentric-Duct Airlift Bioreactors

Abstract: The hydrodynamic environment in bioreactors is an important parameter because cultivation conditions that provide satisfactory oxygen transfer rates tend to be harmful to microorganisms, due to the...

Temporal video scene segmentation using deep-learning

Abstract: The automatic temporal video scene segmentation (also known as video story segmentation) is still an open problem without definite solutions in most cases. Among the available techniques, the ones which shows better results are multimodal using features extracted from multiple modalities. Multimodal fusion may be performed to fuse each modality as a single representation (early fusion) or by each modality segmentation (late fusion), the latter been widely due to multimodal fusion simplicity. Recently, deep learning techniques such as convolutional neural networks (CNN) has been successfully employed to extract features from multiple data sources, easing the development of early fusion methods. However, CNNs cannot adequately learn cues which are temporally distributed along the video due to difficulties to model temporal features data dependencies. A particular deep learning approach which can learn such cues is the recurrent neural network (RNN). Successfully employed on text processing, RNNs are fitted to analyze sequences of data of variable length and may better grasp the temporal relationship among low-level features of video segments, hopefully obtaining more accurate scene boundary detection. This paper goes beyond direct applying RNNs and proposes a new multimodal approach to temporally segment a video into scenes. This approach builds a new architecture carefully combining CNN and RNN capabilities, obtaining better efficacy results on the task when compared with related techniques on a public video dataset.

Microstructural Evolution of Composite 8 WC-(Co, Ni): Effect of the Addition of SiC

Abstract: Tungsten carbide (WC) based cemented carbides, also called hardmetals, are a family of composite materials consisting of carbide ceramic particles embedded in a metallic binder. They are classified as metal matrix composites (MMCs) because the metallic binder is the matrix that holds the bulk material together [1]. WC based composites are used in applications where a good combination of hardness and toughness are necessary [2]. It is usual to add more components to tailor the microstructure of the WC-(Co, Ni) system. The hardness for the cemented carbides based on nickel, increases significantly because of the addition of reinforcements like SiC nanowhisker [3]. In this work, the SiC was considered as an additional component for the composite WC-8(Co, Ni). Four mixtures were prepared with SiC contents ranging from 0 to 3.0 wt%. These mixtures were pressed (200 MPa) and green samples with 25.2 mm of diameter and 40 g were produced. Sintering was carried out in Sinter-HIP furnace (20 bar). Two sintering temperatures were investigated, i.e. 1380 and 1420oC, and the sintering time considered was 60 minutes. The relative density, hardness, linear and volumetric shrinkage were determined. Microstructural evaluation was investigated by optical microscopy and scanning electron microscopy (SEM-FEG). The results showed that the addition of SiC promoted higher densification and grain size growth. The hardness was higher for samples with SiC, so solid solution hardening of the binder was more effective than WC grain size growth.