Automatic Schaeffer's gestures recognition system

Schaeffer's sign language consists of a reduced set of gestures designed to help children with autism or cognitive learning disabilities to develop adequate communication skills. Our automatic recognition system for Schaeffer's gesture language uses the information provided by an RGB‐D camera to capture body motion and recognize gestures using dynamic time warping combined with k‐nearest neighbors methods. The learning process is reinforced by the interaction with the proposed system that accelerates learning itself thus helping both children and educators. To demonstrate the validity of the system, a set of qualitative experiments with children were carried out. As a result, a system which is able to recognize a subset of 11 gestures of Schaeffer's sign language online was achieved.     

Incremental Learning of Skills in a Task-Parameterized Gaussian Mixture Model

Programming by demonstration techniques facilitate the programming of robots. Some of them allow the generalization of tasks through parameters, although they require new training when trajectories different from the ones used to estimate the model need to be added. One of the ways to re-train a robot is by incremental learning, which supplies additional information of the task and does not require teaching the whole task again. The present study proposes three techniques to add trajectories to a previously estimated task-parameterized Gaussian mixture model. The first technique estimates a new model by accumulating the new trajectory and the set of trajectories generated using the previous model. The second technique permits adding to the parameters of the existent model those obtained for the new trajectories. The third one updates the model parameters by running a modified version of the Expectation-Maximization algorithm, with the information of the new trajectories. The techniques were evaluated in a simulated task and a real one, and they showed better performance than that of the existent model.     

Efficient and robust deep learning with Correntropy-induced loss function

Deep learning systems aim at using hierarchical models to learning high-level features from low-level features. The progress in deep learning is great in recent years. The robustness of the learning systems with deep architectures is however rarely studied and needs further investigation. In particular, the mean square error (MSE), a commonly used optimization cost function in deep learning, is rather sensitive to outliers (or impulsive noises). Robust methods are needed to improve the learning performance and immunize the harmful influences caused by outliers which are pervasive in real-world data. In this paper, we propose an efficient and robust deep learning model based on stacked auto-encoders and Correntropy-induced loss function (CLF), called CLF-based stacked auto-encoders (CSAE). CLF as a nonlinear measure of similarity is robust to outliers and can approximate different norms (from \(l_0\) to \(l_2\)) of data. Essentially, CLF is an MSE in reproducing kernel Hilbert space. Different from conventional stacked auto-encoders, which use, in general, the MSE as the reconstruction loss and KL divergence as the sparsity penalty term, the reconstruction loss and sparsity penalty term in CSAE are both built with CLF. The fine-tuning procedure in CSAE is also based on CLF, which can further enhance the learning performance. The excellent and robust performance of the proposed model is confirmed by simulation experiments on MNIST benchmark dataset.     

The other side of ranking schemes: generating weights for specified outcomes

This paper treats the problem of how to determine weights in a ranking, which will cause a selected entity to attain the highest possible position. We establish that there are two types of entities in a ranking scheme: those which can be ranked as number one and those which cannot. These two types of entities can be identified using the “ranking hull” of the data; a polyhedral set that envelops the data. Only entities with data points on the boundary of this hull can attain the number one position. There are no weights that will make an entity whose data point is in the interior of the hull to ever attain the number one position. We deal with these two types of entities separately. In the first case, we propose an approach for finding a set of weights that, under special conditions, will result in a selected entity achieving the top of the ranking without ties and without ignoring any of the attributes. For the second category of entities, we devise a procedure to guarantee that these entities will attain their highest possible position in the ranking. The first case will require using interior point methods to solve a linear program (LP). The second case involves a binary mixed integer formulation. These two mathematical programs were tested on data from a well‐known university ranking.     

Computerized design of advanced straight and skew bevel gears produced by precision forging

The computerized design of advanced straight and skew bevel gears produced by precision forging is proposed. Modifications of the tooth surfaces of one of the members of the gear set are proposed in order to localize the bearing contact and predesign a favorable function of transmission errors. The proposed modifications of the tooth surfaces will be computed by using a modified imaginary crown-gear and applied in manufacturing through the use of the proper die geometry. The geometry of the die is obtained for each member of the gear set from their theoretical geometry obtained considering its generation by the corresponding imaginary crown-gear. Two types of surface modification, whole and partial crowning, are investigated in order to get the more effective way of surface modification of skew and straight bevel gears. A favorable function of transmission errors is predesigned to allow low levels of noise and vibration of the gear drive. Numerical examples of design of both skew and straight bevel gear drives are included to illustrate the advantages of the proposed geometry.     

Semantic-based authorization architecture for Grid

There are a few issues that still need to be covered regarding security in the Grid area. One of them is authorization where there exist good solutions to define, manage and enforce authorization policies in Grid scenarios. However, these solutions usually do not provide Grid administrators with semantic-aware components closer to the particular Grid domain and easing different administration tasks such as conflict detection or resolution. This paper defines a proposal based on Semantic Web to define, manage and enforce security policies in a Grid scenario. These policies are defined by means of semantic-aware rules which help the administrator to create higher-level definitions with more expressiveness. These rules also permit performing added-value tasks such as conflict detection and resolution, which can be of interest in medium and large scale scenarios where different administrators define the authorization rules that should be followed before accessing a resource in the Grid. The proposed solution has been also tested providing some reasonable response times in the authorization decision process.     

A SAT Solver Using Reconfigurable Hardware and Virtual Logic

In this paper, we present the architecture of a new SAT solver using reconfigurable logic and a virtual logic scheme. Our main contributions include new forms of massive fine-grain parallelism, structured design techniques based on iterative logic arrays that reduce compilation times from hours to minutes, and a decomposition technique that creates independent subproblems that may be concurrently solved by unconnected FPGAs. The decomposition technique is the basis of the virtual logic scheme, since it allows solving problems that exceed the hardware capacity. Our architecture is easily scalable. Our results show several orders of magnitude speedup compared with a state-of-the-art software implementation, and also with respect to prior SAT solvers using reconfigurable hardware.     

Superpipelined high-performance optical-flow computation architecture

Optical-flow computation is a well-known technique and there are important fields in which the application of this visual modality commands high interest. Nevertheless, most real-world applications require real-time processing, an issue which has only recently been addressed. Most real-time systems described to date use basic models which limit their applicability to generic tasks, especially when fast motion is presented or when subpixel motion resolution is required. Therefore, instead of implementing a complex optical-flow approach, we describe here a very high-frame-rate optical-flow processing system. Recent advances in image sensor technology make it possible nowadays to use high-frame-rate sensors to properly sample fast motion (i.e. as a low-motion scene), which makes a gradient-based approach one of the best options in terms of accuracy and consumption of resources for any real-time implementation. Taking advantage of the regular data flow of this kind of algorithm, our approach implements a novel superpipelined, fully parallelized architecture for optical-flow processing. The system is fully working and is organized into more than 70 pipeline stages, which achieve a data throughput of one pixel per clock cycle. This computing scheme is well suited to FPGA technology and VLSI implementation. The developed customized DSP architecture is capable of processing up to 170 frames per second at a resolution of 800 × 600 pixels. We discuss the advantages of high-frame-rate processing and justify the optical-flow model chosen for the implementation. We analyze this architecture, measure the system resource requirements using FPGA devices and finally evaluate the system’s performance and compare it with other approaches described in the literature.