Hand gesture recognition system based in computer vision and machine learning

"Lecture notes in computational vision and biomechanics series, ISSN 2212-9391, vol. 19"

Vision-based portuguese sign language recognition system

Vision-based hand gesture recognition is an area of active current research in computer vision and machine learning. Being a natural way of human interaction, it is an area where many researchers are working on, with the goal of making human computer interaction (HCI) easier and natural, without the need for any extra devices. So, the primary goal of gesture recognition research is to create systems, which can identify specific human gestures and use them, for example, to convey information. For that, vision-based hand gesture interfaces require fast and extremely robust hand detection, and gesture recognition in real time. Hand gestures are a powerful human communication modality with lots of potential applications and in this context we have sign language recognition, the communication method of deaf people. Sign lan- guages are not standard and universal and the grammars differ from country to coun- try. In this paper, a real-time system able to interpret the Portuguese Sign Language is presented and described. Experiments showed that the system was able to reliably recognize the vowels in real-time, with an accuracy of 99.4% with one dataset of fea- tures and an accuracy of 99.6% with a second dataset of features. Although the im- plemented solution was only trained to recognize the vowels, it is easily extended to recognize the rest of the alphabet, being a solid foundation for the development of any vision-based sign language recognition user interface system.

Interaction paradigms versus age-related user profiles: an evaluation on content selection

Novel input modalities such as touch, tangibles or gestures try to exploit human's innate skills rather than imposing new learning processes. However, despite the recent boom of different natural interaction paradigms, it hasn't been systematically evaluated how these interfaces influence a user's performance or whether each interface could be more or less appropriate when it comes to: 1) different age groups; and 2) different basic operations, as data selection, insertion or manipulation. This work presents the first step of an exploratory evaluation about whether or not the users' performance is indeed influenced by the different interfaces. The key point is to understand how different interaction paradigms affect specific target-audiences (children, adults and older adults) when dealing with a selection task. 60 participants took part in this study to assess how different interfaces may influence the interaction of specific groups of users with regard to their age. Four input modalities were used to perform a selection task and the methodology was based on usability testing (speed, accuracy and user preference). The study suggests a statistically significant difference between mean selection times for each group of users, and also raises new issues regarding the “old” mouse input versus the “new” input modalities.

User expectations for media sharing practices in open display networks

Open Display Networks have the potential to allow many content creators to publish their media to an open-ended set of screen displays. However, this raises the issue of how to match that content to the right displays. In this study, we aim to understand how the perceived utility of particular media sharing scenarios is affected by three independent variables, more specifically: (a) the locativeness of the content being shared; (b) how personal that content is and (c) the scope in which it is being shared. To assess these effects, we composed a set of 24 media sharing scenarios embedded with different treatments of our three independent variables. We then asked 100 participants to express their perception of the relevance of those scenarios. The results suggest a clear preference for scenarios where content is both local and directly related to the person that is publishing it. This is in stark contrast to the types of content that are commonly found in public displays, and confirms the opportunity that open displays networks may represent a new media for self-expression. This novel understanding may inform the design of new publication paradigms that will enable people to share media across the display networks.

An interactive web-based tool for breast reduction surgery simulation

In this work, we present a 3D web-based interactive tool for numerical modeling and simulation approach to breast reduction surgery simulation, to assist surgeons in planning all aspects related to breast reduction surgery before the actual procedure takes place, thereby avoiding unnecessary risks. In particular, it allows the modeling of the initial breast geometry, the definition of all aspects related to the surgery and the visualization of the post-surgery breast shape in a realistic environment.

Flexible gastrointestinal motility pressure sensors based on aluminum thin-film strain-gauge arrays

This paper reports on an innovative approach to measuring intraluminal pressure in the upper gastrointestinal (GI) tract, especially monitoring GI motility and peristaltic movements. The proposed approach relies on thin-film aluminum strain gauges deposited on top of a Kapton membrane, which in turn lies on top of an SU-8 diaphragm-like structure. This structure enables the Kapton membrane to bend when pressure is applied, thereby affecting the strain gauges and effectively changing their electrical resistance. The sensor, with an area of 3.4 mm2, is fabricated using photolithography and standard microfabrication techniques (wet etching). It features a linear response (R2 = 0.9987) and an overall sensitivity of 2.6 mV mmHg−1. Additionally, its topology allows a high integration capability. The strain gauges’ responses to pressure were studied and the fabrication process optimized to achieve high sensitivity, linearity, and reproducibility. The sequential acquisition of the different signals is carried out by a microcontroller, with a 10-bit ADC and a sample rate of 250 Hz. The pressure signals are then presented in a user-friendly interface, developed using the Integrated Development Environment software, QtCreator IDE, for better visualization by physicians.

Development of a data acquisition system to characterise sustainable energy technologies and to support Web-Based learning (DAQ-WBL)

Driven by concerns about rising energy costs, security of supply and climate change a new wave of Sustainable Energy Technologies (SET’s) have been embraced by the Irish consumer. Such systems as solar collectors, heat pumps and biomass boilers have become common due to government backed financial incentives and revisions of the building regulations. However, there is a deficit of knowledge and understanding of how these technologies operate and perform under Ireland’s maritime climate. This AQ-WBL project was designed to address both these needs by developing a Data Acquisition (DAQ) system to monitor the performance of such technologies and a web-based learning environment to disseminate performance characteristics and supplementary information about these systems. A DAQ system consisting of 108 sensors was developed as part of Galway-Mayo Institute of Technology’s (GMIT’s) Centre for the Integration of Sustainable EnergyTechnologies (CiSET) in an effort to benchmark the performance of solar thermal collectors and Ground Source Heat Pumps (GSHP’s) under Irish maritime climate, research new methods of integrating these systems within the built environment and raise awareness of SET’s. It has operated reliably for over 2 years and has acquired over 25 million data points. Raising awareness of these SET’s is carried out through the dissemination of the performance data through an online learning environment. A learning environment was created to provide different user groups with a basic understanding of a SET’s with the support of performance data, through a novel 5 step learning process and two examples were developed for the solar thermal collectors and the weather station which can be viewed at http://www.kdp 1 .aquaculture.ie/index.aspx. This online learning environment has been demonstrated to and well received by different groups of GMIT’s undergraduate students and plans have been made to develop it further to support education, awareness, research and regional development.

Data-based priors for vector autoregressions with drifting coefficients

This paper proposes full-Bayes priors for time-varying parameter vector autoregressions (TVP-VARs) which are more robust and objective than existing choices proposed in the literature. We formulate the priors in a way that they allow for straightforward posterior computation, they require minimal input by the user, and they result in shrinkage posterior representations, thus, making them appropriate for models of large dimensions. A comprehensive forecasting exercise involving TVP-VARs of different dimensions establishes the usefulness of the proposed approach.

Semi-supervised segmentation of ultrasound images based on patch representation and continuous min cut.

Ultrasound segmentation is a challenging problem due to the inherent speckle and some artifacts like shadows, attenuation and signal dropout. Existing methods need to include strong priors like shape priors or analytical intensity models to succeed in the segmentation. However, such priors tend to limit these methods to a specific target or imaging settings, and they are not always applicable to pathological cases. This work introduces a semi-supervised segmentation framework for ultrasound imaging that alleviates the limitation of fully automatic segmentation, that is, it is applicable to any kind of target and imaging settings. Our methodology uses a graph of image patches to represent the ultrasound image and user-assisted initialization with labels, which acts as soft priors. The segmentation problem is formulated as a continuous minimum cut problem and solved with an efficient optimization algorithm. We validate our segmentation framework on clinical ultrasound imaging (prostate, fetus, and tumors of the liver and eye). We obtain high similarity agreement with the ground truth provided by medical expert delineations in all applications (94% DICE values in average) and the proposed algorithm performs favorably with the literature.

ARCHITECTING USER-CENTRIC PRIVACY-AS-A-SET-OF-SERVICES: DIGITAL IDENTITY RELATED PRIVACY FRAMEWORK

AbstractDigitalization gives to the Internet the power by allowing several virtual representations of reality, including that of identity. We leave an increasingly digital footprint in cyberspace and this situation puts our identity at high risks. Privacy is a right and fundamental social value that could play a key role as a medium to secure digital identities. Identity functionality is increasingly delivered as sets of services, rather than monolithic applications. So, an identity layer in which identity and privacy management services are loosely coupled, publicly hosted and available to on-demand calls could be more realistic and an acceptable situation. Identity and privacy should be interoperable and distributed through the adoption of service-orientation and implementation based on open standards (technical interoperability). Ihe objective of this project is to provide a way to implement interoperable user-centric digital identity-related privacy to respond to the need of distributed nature of federated identity systems. It is recognized that technical initiatives, emerging standards and protocols are not enough to guarantee resolution for the concerns surrounding a multi-facets and complex issue of identity and privacy. For this reason they should be apprehended within a global perspective through an integrated and a multidisciplinary approach. The approach dictates that privacy law, policies, regulations and technologies are to be crafted together from the start, rather than attaching it to digital identity after the fact. Thus, we draw Digital Identity-Related Privacy (DigldeRP) requirements from global, domestic and business-specific privacy policies. The requirements take shape of business interoperability. We suggest a layered implementation framework (DigldeRP framework) in accordance to model-driven architecture (MDA) approach that would help organizations' security team to turn business interoperability into technical interoperability in the form of a set of services that could accommodate Service-Oriented Architecture (SOA): Privacy-as-a-set-of- services (PaaSS) system. DigldeRP Framework will serve as a basis for vital understanding between business management and technical managers on digital identity related privacy initiatives. The layered DigldeRP framework presents five practical layers as an ordered sequence as a basis of DigldeRP project roadmap, however, in practice, there is an iterative process to assure that each layer supports effectively and enforces requirements of the adjacent ones. Each layer is composed by a set of blocks, which determine a roadmap that security team could follow to successfully implement PaaSS. Several blocks' descriptions are based on OMG SoaML modeling language and BPMN processes description. We identified, designed and implemented seven services that form PaaSS and described their consumption. PaaSS Java QEE project), WSDL, and XSD codes are given and explained.

Coating carbon nanotubes with a polystyrene-based polymer protects against pulmonary toxicity.

BACKGROUND: carbon nanotubes (CNT) can have adverse effects on health. Therefore, minimizing the risk associated with CNT exposure is of crucial importance. The aim of this work was to evaluate if coating multi-walled CNT (MWCNT) with polymers could modify their toxicity, thus representing a useful strategy to decrease adverse health effects of CNT. We used industrially-produced MWCNT uncoated (NT1) or coated (50/50 wt%) with acid-based (NT2) or polystyrene-based (NT3) polymer, and exposed murine macrophages (RAW 264.7 cell line) or Balb/c mice by intratracheal administration. Biological experiments were performed both in vitro and in vivo, examining time- and dose-dependent effects of CNT, in terms of cytotoxicity, expression of genes and proteins related to oxidative stress, inflammation and tissue remodeling, cell and lung tissue morphology (optical and transmission electron microscopy), and bronchoalveolar lavage fluid content analysis.RESULTS: extensive physico-chemical characterization of MWCNT was performed, and showed, although similar dimensions for the 3 MWCNT, a much smaller specific surface area for NT2 and NT3 as compared to NT1 (54.1, 34 and 227.54 m(2)/g respectively), along with different surface characteristics. MWCNT-induced cytotoxicity, oxidative stress, and inflammation were increased by acid-based and decreased by polystyrene-based polymer coating both in vitro in murine macrophages and in vivo in lung of mice monitored for 6 months.CONCLUSIONS: these results demonstrate that coating CNT with polymers, without affecting their intrinsic structure, may constitute a useful strategy for decreasing CNT toxicity, and may hold promise for improving occupational safety and that of general the user.

Open to the users' needs : combining user-centered design, standards and open source software

This case study introduces our continuous work to enhance the virtual classroom in order to provide faculty and students with an environment open to their needs, compliant with learning standards and, therefore compatible with other e-learning environments, and based on open source software. The result is a modulable, sustainable and interoperable learning environment that can be adapted to different teaching and learning situations by incorporating the LMS integrated tools as well as wikis, blogs, forums and Moodle activities among others.

Evaluation of an online, case-based interactive approach to teaching pathophysiology

AIM: The aim of this study was to evaluate a new pedagogical approach in teaching fluid, electrolyte and acid-base pathophysiology in undergraduate students. METHODS: This approach comprises traditional lectures, the study of clinical cases on the web and a final interactive discussion of these cases in the classroom. When on the web, the students are asked to select laboratory tests that seem most appropriate to understand the pathophysiological condition underlying the clinical case. The percentage of students having chosen a given test is made available to the teacher who uses it in an interactive session to stimulate discussion with the whole class of students. The same teacher used the same case studies during 2 consecutive years during the third year of the curriculum. RESULTS: The majority of students answered the questions on the web as requested and evaluated positively their experience with this form of teaching and learning. CONCLUSIONS: Complementing traditional lectures with online case-based studies and interactive group discussions represents, therefore, a simple means to promote the learning and the understanding of complex pathophysiological mechanisms. This simple problem-based approach to teaching and learning may be implemented to cover all fields of medicine.

Curriculum sequencing for an e-learning system based on learning styles

This work shows the use of adaptation techniques involved in an e-learning system that considers students' learning styles and students' knowledge states. The mentioned e-learning system is built on a multiagent framework designed to examine opportunities to improve the teaching and to motivate the students to learn what they want in a user-friendly and assisted environment

Kernel-based methods for change detection in remote sensing images

Nowadays, the joint exploitation of images acquired daily by remote sensing instruments and of images available from archives allows a detailed monitoring of the transitions occurring at the surface of the Earth. These modifications of the land cover generate spectral discrepancies that can be detected via the analysis of remote sensing images. Independently from the origin of the images and of type of surface change, a correct processing of such data implies the adoption of flexible, robust and possibly nonlinear method, to correctly account for the complex statistical relationships characterizing the pixels of the images. This Thesis deals with the development and the application of advanced statistical methods for multi-temporal optical remote sensing image processing tasks. Three different families of machine learning models have been explored and fundamental solutions for change detection problems are provided. In the first part, change detection with user supervision has been considered. In a first application, a nonlinear classifier has been applied with the intent of precisely delineating flooded regions from a pair of images. In a second case study, the spatial context of each pixel has been injected into another nonlinear classifier to obtain a precise mapping of new urban structures. In both cases, the user provides the classifier with examples of what he believes has changed or not. In the second part, a completely automatic and unsupervised method for precise binary detection of changes has been proposed. The technique allows a very accurate mapping without any user intervention, resulting particularly useful when readiness and reaction times of the system are a crucial constraint. In the third, the problem of statistical distributions shifting between acquisitions is studied. Two approaches to transform the couple of bi-temporal images and reduce their differences unrelated to changes in land cover are studied. The methods align the distributions of the images, so that the pixel-wise comparison could be carried out with higher accuracy. Furthermore, the second method can deal with images from different sensors, no matter the dimensionality of the data nor the spectral information content. This opens the doors to possible solutions for a crucial problem in the field: detecting changes when the images have been acquired by two different sensors.