Daniel Moraes

Design and Implementation Issues for Modern Remote Laboratories

The design and implementation of remote laboratories present different levels of complexity according to the nature of the equipments operated by the remote laboratory, the requirements imposed on the accessing computers, the network linking the user to the laboratory, and the type of experiments the laboratory supports. This paper addresses the design and implementation of remote laboratories employing web technologies, both at the client and the server side. These types of remote laboratories are called WebLabs, and can be deployed over different networks such as the public internet, campuswide networks, or high-speed private networks. Although most published works on WebLabs focus on their functional and operational aspects, nonfunctional requirements related to security, quality of service, and federated operation of WebLabs have received little attention. This paper addresses how these requirements can be incorporated into WebLab design, and discusses the most appropriate web technologies to fulfill such requirements.

Video pornography detection through deep learning techniques and motion information

Recent literature has explored automated pornographic detection - a bold move to replace humans in the tedious task of moderating online content. Unfortunately, on scenes with high skin exposure, such as people sunbathing and wrestling, the state of the art can have many false alarms. This paper is based on the premise that incorporating motion information in the models can alleviate the problem of mapping skin exposure to pornographic content, and advances the bar on automated pornography detection with the use of motion information and deep learning architectures. Deep Learning, especially in the form of Convolutional Neural Networks, have striking results on computer vision, but their potential for pornography detection is yet to be fully explored through the use of motion information. We propose novel ways for combining static (picture) and dynamic (motion) information using optical flow and MPEG motion vectors. We show that both methods provide equivalent accuracies, but that MPEG motion vectors allow a more efficient implementation. The best proposed method yields a classification accuracy of 97.9% - an error reduction of 64.4% when compared to the state of the art - on a dataset of 800 challenging test cases. Finally, we present and discuss results on a larger, and more challenging, dataset.

Pornography classification: The hidden clues in video space-time.

As web technologies and social networks become part of the general publics life, the problem of automatically detecting pornography is into every parents mind - nobody feels completely safe when their children go online. In this paper, we focus on video-pornography classification, a hard problem in which traditional methods often employ still-image techniques - labeling frames individually prior to a global decision. Frame-based approaches, however, ignore significant cogent information brought by motion. Here, we introduce a space-temporal interest point detector and descriptor called Temporal Robust Features (TRoF). TRoF was custom-tailored for efficient (low processing time and memory footprint) and effective (high classification accuracy and low false negative rate) motion description, particularly suited to the task at hand. We aggregate local information extracted by TRoF into a mid-level representation using Fisher Vectors, the state-of-the-art model of Bags of Visual Words (BoVW). We evaluate our original strategy, contrasting it both to commercial pornography detection solutions, and to BoVW solutions based upon other space-temporal features from the scientific literature. The performance is assessed using the Pornography-2k dataset, a new challenging pornographic benchmark, comprising 2000 web videos and 140h of video footage. The dataset is also a contribution of this work and is very assorted, including both professional and amateur content, and it depicts several genres of pornography, from cartoon to live action, with diverse behavior and ethnicity. The best approach, based on a dense application of TRoF, yields a classification error reduction of almost 79% when compared to the best commercial classifier. A sparse description relying on TRoF detector is also noteworthy, for yielding a classification error reduction of over 69%, with 19× less memory footprint than the dense solution, and yet can also be implemented to meet real-time requirements.

A network architecture for large mobile robotics environments

Mobile robotics environments must adopt networking solutions that provide secure and reliable communications for the mobile robots across wide areas such as hospitals, factories, farms, etc. This paper proposes a network architecture for large mobile robotic environments built above the existing networking infrastructures. The architecture builds an overlay network above the already deployed network. The overlay network must fulfill the requirements demanded by mobile robotic applications, mainly, communication continuity during handover, security, and quality of service. A prototype of this architecture was implemented and evaluated in a mobile robotic environment composed of Pioneer P3-DX mobile robots accessed through the Internet. Results from simulation show that the architecture scales well in larger networking scenarios.

A reference architecture for micro-mobility support in IP networks

This paper presents an architecture for supporting micro-mobility in IP networks. The architecture is presented in terms of its functional blocks and interaction protocols. The architecture, named mobility plane architecture (MPA), offers a network-centric solution for micro-mobility. This means that all the functions necessary for supporting Layer-3 mobility are placed on the network, not on the mobile node. The architecture does not interfere with services already deployed in the transport network, even services supporting macro mobility such as Mobile IP. Another key point is the total independence of the architecture from IPv6, meaning that it can be deployed on existing IPv4 networks.

MPA: A Network-Centric Architecture for Micro-Mobility Support in IP and MPLS Networks

Micro-mobility protocols aim to improve localized mobility by reducing the handover overheads. In this paper we present the Mobility Plane Architecture (MPA). This architecture was designed to support micro-mobility in standard IP or MPLS/GMPLS networks in a network-centric way, that is, the burden demanded by micro-mobility is placed on the network, not on the mobile nodes. The main advantages of this architecture are its independence of IPv6 and the absence of new protocols for supporting L3 mobility.

Low false positive learning with support vector machines

Novel 2-level classification method for low false positive classification.Level 1 defines a decision boundary through an SVM classifier.Level 2 defines a sensitive area around the decision boundary.The sensitive area are analyzed by a second classifier to control false positives.Methods effectiveness showed trough comparisons to other solutions in 33 datasets. Most machine learning systems for binary classification are trained using algorithms that maximize the accuracy and assume that false positives and false negatives are equally bad. However, in many applications, these two types of errors may have very different costs. In this paper, we consider the problem of controlling the false positive rate on SVMs, since its traditional formulation does not offer such assurance. To solve this problem, we define a feature space sensitive area, where the probability of having false positives is higher, and use a second classifier (unanimity k-NN) in this area to better filter errors and improve the decision-making process. We call this method Risk Area SVM (RA-SVM). We compare the RA-SVM to other state-of-the-art methods for low false positive classification using 33 standard datasets in the literature. The solution we propose shows better performance in the vast majority of the cases using the standard Neyman-Pearson measure.

Design and Implementation of a Network-Centric Micro-Mobility Architecture

This paper presents the design and implementation of the Mobility Plane Architecture (MPA). MPA is a network architecture that provides micro-mobility in a network-centric way, that is, the burden of supporting micro-mobility is placed on the network and not on the mobile nodes. The implementation employs RSVP-TE to establish IP/IP and MPLS point-to-multipoint tunnels in order to direct traffic to the mobile nodes. DHCP is used for tracking the mobile node locations and RSVP-TE opaque data carries the location information to install routes to them. Results obtained in a lab-sized network and some quantitative comparisons between MPA and other related work are presented as well.

Temporal Robust Features for Violence Detection

Automatically detecting violence in videos is paramount for enforcing the law and providing the society with better policies for safer public places. In addition, it may be essential for protecting minors from accessing inappropriate contents on-line, and for helping parents choose suitable movie titles for their children. However, this is an open problem as the very definition of violence is subjective and may vary from one society to another. Detecting such nuances from video footages with no human supervision is very challenging. Clearly, when designing a computer-aided solution to this problem, we need to think of efficient (quickly harness large troves of data) and effective detection methods (robustly filter what needs special attention and further analysis). In this vein, we explore a content description method for violence detection founded upon temporal robust features that quickly grasp video sequences, automatically classifying violent videos. The used method also holds promise for fast and effective classification of other recognition tasks (e.g., pornography and other inappropriate material). When compared to more complex counterparts for violence detection, the method shows similar classification quality while being several times more efficient in terms of runtime and memory footprint.

A Network Architecture for Providing Micro-Mobility in MPLS/GMPLS Networks

The Mobile Plane Architecture (MPA) is a network architecture that provides micro-mobility in a network-centric way, that is, the burden of supporting micro-mobility is placed on the network and not on the mobile nodes. MPA employs an overlay network above an IPv4 or IPv6 transport network in order to direct traffic to the mobile nodes. The overlay network is built by establishing point-to-multipoint (P2MP) tunnels through the transport network. This paper presents an implementation of MPA over MPLS/GMPLS networks. It is shown that the combination of P2MP tunneling with label stacking is an effective way for providing mobility services in MPLS/GMPLS transport networks. Results obtained in a lab-sized network are presented as well.