Search

Augmented, mixed and virtual reality are changing the way people interact and communicate. Five dimensional communications and services, integrating information from all human senses are expected to emerge, together with holographic communications (HC), providing a truly immersive experience. HC presents a lot of challenges in terms of data gathering and transmission, demanding Artificial Intelligence empowered communication technologies such as 5G. The goal of the paper is to present a model of a context-aware holographic architecture for real time communication based on semantic knowledge extraction. This architecture will require analyzing, combining and developing methods and algorithms for: 3D human body model acquisition; semantic knowledge extraction with deep neural networks to predict human behaviour; analysis of biometric modalities; context-aware optimization of network resource allocation for the purpose of creating a multi-party, from-capturing-to-rendering HC framework. We illustrate its practical deployment in a scenario that can open new opportunities in user experience and business model innovation.

With the advent of Sixth Generation (6G) telecommunication systems already envisioned, increased effort is made to further develop current communication technologies, so they can be incorporated together with the novel ones, to deliver uninterrupted and satisfactory service for any application in every location on the ground, underwater, in the air, or in space. One such technology is Cognitive Radio (CR) which has received much attention due to its potential for increase of utilization, especially in the bands below 6 GHz. The main enabler for CR is spectrum sensing because it provides the opportunity for dynamic assessment of the radio environment to identify unused channels. This functionality has been the object of many research works for that very reason. In spite of this, the provision of accurate and fast spectrum characterization in time, frequency and space has proven to be a non-trivial task. This paper presents a detailed review of probabilistic spectrum sensing methods classified by the feature they extract from the received signal samples, to provide accurate detection of the primary user (PU) signal. The main design characteristics (such as probability of detection, robustness to noise and fading, signal/noise model assumptions, and computational complexity), strengths and weaknesses for each type are also summarized. Based on current concepts for 6G networks and applications, a framework for human-centric cognition-based wireless access is presented, which specifies the role of spectrum sensing-based CR in future networks.

Many Research Institutions are increasingly focusing on the efficient transfer of Intellectual Property to industry and businesses, as well as governments and society. Successful and long-term Knowledge Transfer policies can improve collaboration between research institutions, businesses, and the public sector. The revenue generated by licensing technologies and developing spin-off companies can help research organizations become more financially sustainable. This study focuses on knowledge transfer in research institutes, with a use case scenario in the HOLOTWIN project as an example.

Recognizing specific actions, activities and goals of an individual in any environment (constraint or unconstrained) will be a key feature for holographic communications in order to achieve a sense of reality and naturalness of the face-to-face interaction. While activity and action recognition are tasks that humans perform naturally and with little exertion, they are still a challenge from the point of view of artificial intelligence in the context of deep learning. The main goal of this paper is to present a technique for human activity recognition with semantically guided graph-convolutional network based on auto-regressive moving average (ARMA) filters, for the purpose of holographic communication. The semantic is introduced by the human skeleton representation. By recognizing the activity, we can plan for the next step in the proposed architecture: prediction; thus solving some of the challenges imposed by real time constraints and channel limits when transmitting huge and heterogeneous amounts of data for this type of communication even in 5G.

Analysis on the human motion can reveal patterns proven very useful in human-machine interactions, medical applications and ambient assisted living. One such analysis is human motion prediction consisting of predicting human pose in a set of time instances contained in constrained time window of up to 1 to 2 seconds. This prediction is done by analyzing previous motion, i.e. set of previous poses, within a selected time window. In this paper we propose to predict human motion using Gated Recurrent Unit (GRU) network, a variant of Recurrent Neural Network. The prediction is based on human skeleton model and joints position change in time. We further optimize the GRU by substituting the weighting of inputs and recurrent outputs with convolution utilizing the graph structure of the human skeleton. We validate our proposed network by testing it on publically available dataset and providing state of the art results in comparison with other popular methods.

Globally, there is a huge problem with lack of physical activity, obesity and many diseases due to malnutrition, lack of movement, sedentary lifestyle, etc. Exasperated by the current worldwide situation where almost all of the active population including children of all ages is sequestered at home with no opportunities to do physical exercises outdoors or in a gym, we propose a prototype of a Human Centric Cyber Physical System to measure and analyze physical activities by employing a holographic virtual coach. The system can identify the user from a multi-view camera setup at the begging of each session and a program-based algorithm will detect and analyze his or hers periodic activities. The algorithm can detect the beginning of a series of exercises of a given type, measures the duration of each of them, detects the end of the physical activity sequence, and measures the number of repetitions and deviation from a setup norm. The virtual coach application is managed by the user’s voice and gestures based on Microsoft Hololens.

This paper focuses on researching and identifying strengths and shortcomings of the various educational approaches taken in response to the COVID-19 pandemic and the extent, to which these approaches have been able to ensure effective and equitable access to quality learning opportunities for all. The authors research open initiatives that implement digital education ecosystems, and technologies, which promote digital readiness of educators, educations systems and institutions, at world level and studies and evaluates the benefits of cross-sector collaboration. The prevailing issues and possible solutions for flexible credentialing are outlined and a technological model of a digital space for the exchange of digital education content, methods, practices and tools, at European level is proposed. The relevant EU recovery mechanisms, including the Digital Education Action Plan, Skills Agenda, European Pact for Skills and Funding Frameworks are also presented.

One of the challenges of planetary science is the age determination of geological units on the surface of the different planetary bodies in the solar system. This serves to establish a chronology of the geological events occurring on these different bodies, hence to understand their formation and evolution processes. An approach for dating planetary surfaces relies on the analysis of the impact crater densities with size. Approaches have been proposed to automatically detect impact craters in order to facilitate the dating process. They rely on color values from images or elevation values from Digital Elevation Models (DEM). In this article, we propose a new approach for crater detection, more specifically using their rims. The craters can be characterized by a round shape that can be used as a feature. The developed method is based on an analysis of the DEM geometry, represented as a 3D mesh, followed by curvature analysis. The classification process is done with one layer perceptron. The validation of the method is performed on DEMs of Mars, acquired by a laser altimeter aboard NASA’s Mars Global Surveyor spacecraft and combined with a database of manually identified craters. The results show that the proposed approach significantly reduces the number of false negatives compared to others based on topographic information only.

In recent years, research related to the analysis of human activity has been the subject of increased attention by engineers dealing with computer vision, and particularly that which utilizes deep learning. In this paper, we propose a method for classification of human activities, composed of 3D skeleton data. This data is normalized beforehand and represented in two forms, which are fed to a neural network with parallel convolutional and dense layers. After the network is trained, the training data is propagated again to infer the output from the second last layer. This output is used for training a Support Vector Machine. All hyperparameters were found using the Bayesian Optimization strategy on the PKU-MMD dataset. Our method was tested on the UTD-MHAD dataset, achieving an accuracy of 92.4%

Brain-Computer Interfaces (BCIs) are an approach that enables humans to interact with their surroundings by brain generated control signals. Electroencephalographic (EEG) signals that records electrical activity through the scalp might contain superfluous artifacts suppressing some valuable information. Thus the EEG signal denoising is an important stage of the EEG data analyses. The paper presents an experimental comparison of several classification approaches for 2-class motor imagery EEG data classification and explores the influence of wavelet signal denoising on the classification accuracy.

Social distancing is not just about keeping away from loved ones, shutting schools or working from home. A strict physical separation must be maintained in order to prevent the spread of COVID-19. Yet, front-line workers in healthcare, public safety, retail, manufacturing and transportation and logistics have found it particularly challenging to maintain the 1,5 to 2 meters separation from other people while doing their everyday jobs. In this paper we have presented a simple real-time method for distance estimation with any kind of camera between people applicable either in close quarters or open spaces. The aim is to help easily and non obtrusively alert if the correct distance between people is not maintained. The initial results are promising and the error between the real and the measured values is in statistical margin. At the same time, the proposed system monitors people without facial recognition. The data collected is anonymous and do not require the use of facial recognition or any other personal information.

Ultra-dense Internet of Things (IoT) networks in the unlicensed bands are subject to significant interference due to their high deployment density and attacks from malicious users. The interference has a non-uniform distribution, when examined in 3D space which leads to the need of volumetric monitoring for precise estimation of spectrum utilization. This paper analyzes the indoor spectrum occupancy for two promising IoT standards - LoRa and WiFi, through 3D interference maps and spatial throughput (ST) analysis produced using an automated measurement tool. The system is implemented through software-defined radio (SDR) platforms. Fading impacts the interference distribution not only in a single plane but also in height. In addition, the experimental results show that the received interference power within an area (measured in m2) and a volume (in m3), and consequently the ST, vary dynamically. Thus, the worst-case interference rather the mean value, needs to be considered in realistic dense scenarios for beyond 5G IoT networks.

Activity recognition is a key component of context-aware holographic communication to support optimal quality flow of data, but conventional approaches often lack in semantic information and context-awareness due to problems such as difficulty identifying the activity which the invidiual performs; overfitting when building activity models; collection of a large amount of labeled data from each end user. This paper presents a fully developed multi-view RGB-D system based on user-specific metrics - facial features coupled with a body skeleton. The system employs a skeleton-based approach. We test the performance of the proposed architecture in a controlled environment.

The identification of trees is becoming increasingly important for the protection of plant species and the regulation of timber trade worldwide. Anatomical features such as vessels, fibers, parenchyma and rays play a significant role in the identification of individual trees. The unique cell structure of each of the hardwood species varies widely among the intraspecific species and describes their characteristic features. In this article we propose a method for detection of geometric features in tree samples, such as vessels, through morphological analysis and segmentation.

The planetary surfaces are explored by ground samples or by indirect observations such as images and topographic data. This article examines the topography data from the Mars orbiter laser altimeter (MOLA). The analysis is based on the noise variance of the minimal curvature. A histogram approach to assessing longitudinal and transverse lines laying on geographical features (such as craters) in altimeter data is suggested.

An increasing trend towards making the modulation classification (MC) algorithms better-suited for real world implementation in Cognitive Radio (CR) equipment, can be seen as recent works adopt novel agile deep learning models and datasets which include different kinds of signal impairments. Considering the large range of studies in the field, a unifying examination of the common state-of-the-art methods will benefit the future developments by providing a starting point for comparison of their efficiency for recognition of various communication signals. The purpose of this study is to provide a comparative analysis which gives clarity on the ways in which the dataset’s type and contents (statistical features of the signals as well as the presence of noise) influences the classification performance. Thus, a more systematic approach to choosing the appropriate input data and a suitable classifier model can be employed in future developments.

The ad-hoc network formed by Bluetooth works on radio frequency links. The security aspect of Bluetooth has to be handled more carefully. The radio frequency waves have a characteristic that the waves can pierce the obstructions in the communication path, get rid of the requirement of line of sight between the communicating devices. We propose a software model of man-in-the-middle attack along with unauthorized and authorized transmitter and receiver. Advanced White Gaussian Noise channel is simulated in the designed architecture. The transmitter uses Gaussian Frequency Shift Keying (GFSK) modulation like in Bluetooth. The receiver uses GFSK demodulation. In order to validate the performance of the designed system, bit error rate (BER) measurements are taken with respect to different time intervals. We found that BER drops roughly 18% if hopping duration of 150 seconds is chosen. We propose that a Bluetooth system with hopping rate of 0.006 Hz is used instead of 10Hz.

Signals provided by the ElectroEncephaloGraphy (EEG) are widely used in Brain-Computer Interface (BCI) applications. They can be further analyzed and used for thinking activity recognition. In this paper we proposed an algorithm that is able to recognize five mental tasks using 6 channel EEG data. The main idea is to separate the raw EEG signals into several frames and compute their spectrums. Next, a second-order derivative of Gaussian is applied to extract features and an optimum Gaussian kernel parameters grid search is performed with the help of cross-validation. The extracted features are further reduced by Principal Component Analysis. The processed data is utilized to train SVM classifier which is used for mental tasks recognition afterwards. The performance of the algorithm is estimated on publically available dataset. In terms of 5 folds cross-validation we obtained an average of 82.7% recognition rate (accuracy). Additional experiments were conducted using leave-one-out cross-validation where 67.2% correct classification was reported. Comparison to several state-of-the art methods reveals the advantages of the proposed algorithm.

Virtual reality (VR) active video gaming, also called ”exergaming” has become an emerging trend in entertainment but also fitness, education and health sectors. It requires bodily movements to play and function as a form of physical activity (PA). Since this type of gaming is becoming more and more popular and accessible and the trends suggest that it may become a mainstream media, the scientific community is discussing on its usefulness. There are number of claims about its ability to improve health via an increase in PA but researchers also argue that it is still a form of time spent in front of monitor which comes with many negative sides. The aim of this paper is to do an overview of the scientific research domain and outline the strengths, weaknesses and opportunities of immersive VR gaming to promote PA. The available evidence indicates that active video games can replace sedentary time spent in front of a computer or TV with physical activity and make a measurable contribution to the improvement of health. However whether it is a sustainable way to motivate gamers in PA is questionable.

During the last ten years, Mars has been extensively explored and mapped by several NASA and ESA orbital missions, generating large datasets of high-resolution images. This type of information helps the understanding of impact processes occurring on the surface of celestial bodies. In this work, we introduced a novel automated approach for detection and boundary extraction of Martian impact craters. We implemented a pyramidal image representation and classical morphological operations, involving Hough transform (HT), which identified regions with a specific circular form. This was tested on 3D mesh data of Mars, provided by Mars Orbiter Laser Altimeter (MOLA). We will demonstrate the potential and usefulness of such automated approach in planetary science.

In this paper a comparative analysis is presented among 4 algorithms employing tensor representation of videos for background modelling and subtraction aiming the estimation of fire dispersal. The algorithms are HoRPCA by IALM, Tucker-ALS, CP-ALS, and t-SVD. They are applied over a database of 6 videos containing fires at different stage of spreading, recorded at different scale and angle of perspective. In part of the videos intense smoke is also present. Decomposition times, full processing times with preprocessing stage and accuracy of fire dispersal in terms of relative number of correctly detected pixels forming the flame areas to all flame pixels from the original recordings are registered. Positive results are obtained which reveal the applicability of the tested algorithms for fire dispersal estimation and with the in-depth analysis of experimental results a selection in order of preference could be made for future applications given the circumstances at which fire breaks out.

In most of our daily lives, we observe repetitive (periodic) movements of our surrounding objects. In physiotherapy, for example, during the physical exercises, the analysis of the periodicity in the patient's performance would help to evaluate and monitor the his recovery process. Cyber Physical Systems for smart environments have great potential in providing context-aware, automated support in everyday life of people that need adaptive occupational therapy. We present a method and program-based algorithm for the detection and analysis of periodic activities of subjects performing physical exercises from video sequences. The algorithm detects the beginning of a series of exercises of a given type, measures the duration of each of them, detects the end of the sequence and measures the number of repetitions. Experimental results with publicly available video dataset will prove the applicability of the proposed method for detection of repetitive actions.

The exploration of planetary surfaces lies on images and topographic data, typically derived by laser altimetry from planetary missions. This paper deals with an examination of the topography data from the Mars orbiter laser altimeter (MOLA). MOLA data to be processed is massive and 3D dimensional but the obtained resolution of the martian surface is limited to the altimeter data density, especially along the cross-track directions. Therefor a tensor representation can be used to model these types of noise. In this paper an approach to assessing along-track and crosstrack noise on altimeter data using a tensor decomposition is proposed. Experimental results show that the quality of the reconstituted signal in the different planes are excellent.

Nowadays, the way people see the world and interact between each other is changing; new technologies are paving the way of innovative solutions for communication between humans. The recent research work on relevant topics and the rapid development of powerful hardware and software implementations allow for building of new communications pathways for lifelike interaction between people in different locations called human bond communication. One very new technology may offer the practical application of human bond communication is holoportation, but it also presents a lot of challenges in terms of digital data gathering and transmission. By combining appropriate technics, we propose an approach for holoportation combined with haptic technology between same controlled environments. We present a conceptual model of holoportation architecture for real time communication based on highly accurate 3D modelling of the human face and body, recognition and prediction of human actions and facial expressions to achieve realistic communications. Designed this way the proposed conceptual model of the holoportation system addresses the challenges from the information transmission aspect where real time constraints and narrowband channels are imposed, while big amounts of data, such as 3D body models of humans need to be transmitted.

As the standardization process for the coming 5G networks has entered its final stages, the research activities are ever growing with emphasis on practical deployment scenarios which incorporate many aspects of this family of technologies. Among them are the Ultra-dense network (UDN) and Cognitive Radio (CR) concepts which have both in their own right been objects of intense scientific interest which has yielded many significant accomplishments. Integrating CR functionalities in UDN is a viable solution to the spectrum shortage issue which is ever more critical, especially in such massive deployment scenarios. However, it has only been shallowly explored. This paper expands on the basic approaches for the realization of UDN and CR networks, to present a detailed conceptual scheme for the operation of CR-based nodes operating within the context of UDN along with their relevant functionalities.

Electrocardiogram (ECG) based affective computing is a new research field that aims to find correlates between human emotions and the registered ECG signals. Typically, emotion recognition systems are personalized, i.e. the discrimination models are subject-dependent. Building subject-independent models is a harder problem due to the high ECG variability between individuals. In this paper, we study the potential of two machine learning methods (Logistic Regression and Artificial Neural Network) to discriminate human emotional states across multiple subjects. The users were exposed to movies with different emotional content (neutral, fear, disgust) and their ECG activity was registered. Based on extracted features from the ECG recordings, the three emotional states were partially discriminated.

The ability to develop innovative Business Models (BMs) with high speed and good Return of Investment has become the cornerstone for the competitiveness of many successful businesses operating in many different Business Model Ecosystems (BMES) today. Business Model Innovation (BMI) appears even more important in these days—as well as in the Biogas 2020 BMES—as digitalization affect all areas of innovating and operating businesses. Increased digitalization leads to new ways of working, communicating, cooperating and supporting BMI. Thanks to these new technologies todays businesses are more interconnected in physical, digital and virtual networks. The cooperation between businesses, academia and the use of competences across different domains becomes more important to strengthen BMI and advance BMI. One main challenge however of digitization of BMI is modeling human behavior and cognitive processes in BMI processes into logical expressions that can be digitized and automated. Our state-of-the-art review shows that has been no research in this field related to BMI. The aim is therefore to find a way to value the BMI process and support the BMI process in a more advanced way than previously supported by sensor technologies. This paper report on a first experiment carried out in a cross interdisciplinary CGC research group with 3 different European Biogas Technological BMI (TBMI) challenges. The aim of this first experiment was twofold. (1) To support the test and further development of a conceptual model of a future agile BMI learning environment first presented at the Global Wireless Summit 2017 conference in Cape Town, South Africa. The the aim is later to adapt these results, findings and experiment recommendation to prepare a largescale global experiment in 2019. The used sensor technology backbone was to this experiment made as a special build cloud-based sensing BM and sensing BMI room—imbedded in special built BMI rooms called B-lab/Beecube. The B-labs was embedded with advanced mobile and wireless sensors, both environmental and wearable by the participants. The aim was to test use and utilization of these sensors in BMI to improve observations, improve analyzing abilities and predict in a latter stage human behavior in BMI processes, (2) to support the development of an advanced network based BMI environment and in a latter perspective to support a Multi Business Model Innovation (MBMI) framework embedded with AI, machine learning, AI, Multi Business model archetypes and pattern techniques. In our experiment we were able to use the Biogas 2020 BMES and project as a testbed. We report on our preliminary findings and draw some proposals to how the large scale experiment technical and BMI wise can be adjusted.

In this paper we present an algorithm for providing better user experience in AR applications based on HMD (Head Mounted Display), pair of cameras and Deep Learning semantic segmentation. The user can view the scene, captured by the cameras using HMD along with augmented information regarding distances to all detected objects. The cameras are attached to the HMD and must be located as close as possible each other. Moreover one of the cameras is shifted at a known distance closer to the scene with respect to the other camera. Based on the traditional pinhole camera model and estimating the size of the projection of a given object in pixel coordinate systems in both of the cameras, we are able to calculate the distance from the cameras to the object. For location and calculation of the size of projected objects we used a semantic segmentation based on deep learning algorithm.

Adaptive and interactive mental engagement combined with positive emotional state are requirements for an optimal outcome of the neuro-rehabilitation process for patients with brain damage usually caused by TBI (traumatic brain injury), stroke or brain disease such as cancer, epilepsy, and Alzheimer's disease. We propose a method for automatic pain recognition in video sequences using the landmarks data from Supervised Descent Method and applying Support Vector Machine (SVM) for data classification. This method is suitable for being part of assistive medical system for neuro-rehabilitation of patients with TBI. The experiments with a video dataset with patients with shoulder pain show very good recognition rate (95,7%) for recognizing the painful facial states of the subjects.

Elderly with declining cognitive functions, their family members and assistive care providers have all identified good “quality of life” as a central element in living with diseases such as dementia and mild cognitive impairment. Quality of life concerns mood, sleep, daily activities performance, pleasant engagements, physical mobility and healthy lifestyle. Sleep quality is essential for the human being to maintain good physical and emotional health. And sleep disorders can lead to severe physical and emotional effects, e.g. cognitive decline and mental health complications. So being able to measure and evaluate sleep quality and monitor mood is inseparably linked to every assistive living system. Sensor-based sleep and mood monitoring systems promise to prolong independent living of elderly people with declining physical and cognitive functions. In this paper, an implementation of sleep and mood monitoring and estimation of intelligent modules part of Ambient Assisted Living architecture is presented. A web application displays the computational results by providing information and recommendations on how to improve sleep hygiene to caregivers and coaches the elderly into a healthy sleeping behavior, based on their personal rhythm and health problems.

The knowledge of the origin and development of all bodies in the solar system begins with understanding the geologic history and evolution of the universe. The only approach for dating celestial body surfaces is by the analysis of the crater impact density and size. In order to facilitate this process, automatic approaches have been proposed for the impact craters detection. In this article, we propose a novel approach for detecting craters’ rims. The developed method is based on a study of the Digital Elevation Model (DEM) geometry, represented as a 3D triangulated mesh. We use curvature analysis, in combination with a fast local quantization method to automatically detect the craters’ rims with artificial neural network. The validation of the method is performed on Barlow’s database.

Individuals with Mild Cognitive Impairment (MCI) currently have few treatment options against memory loss. Solutions for caring for the elderly both efficacious and cost-effective are given by Ambient Assisted Living (AAL) architecture, promising the improvement of the Quality of Life (QoL) of patients. QoL factors that are important for the MCI patients include mood, pleasant engagements, physical mobility and health, and the ability to perform activities of daily living. In this paper, we propose a daily activity reasoner that monitors, measures and analyses in real time several everyday events for building habits diary and detecting abnormal behavior of the user, part of an effective AAL system. The proposed solution is based on a combination of mean shift clustering algorithm. The reasoner offers two primary functionalities: habits building and duration and frequency of events. The reasoner can predict the behavior and detect (slow or fast) changes that might indicate modification in the health status of the user.

An average human spends about one third of his life sleeping so quality of sleep is essential for the human being to maintain good physical and emotional health. Sleep disorders may introduce severe physical effects, e.g. cognitive impairments and mental health complications. So being able to measure and evaluate sleep behavior is important for health practitioners and the users themselves. In this paper, we present the implementation of the Sleep Lifestyle Reasoner part of AAL platform which allows detection of minor or major deviations in the sleeping patterns in MCI and COPD patients indicating changes in their health status. The output of the reasoner is fed to the My Sleep Web Application that provides recommendations to improve sleep hygiene and coaches the users into a healthy sleeping behavior, based on their personal rhythms and problems. It also supports the informal caregiver by providing insights on the sleeping behavior of the patient.

Internet of Things (IoT) is the next wave of technology in the society where we are visualizing automation in almost everything except the human brain. In this environment, various nodes will collect the data and each node may use a wireless communication technology like Bluetooth, Wireless Fidelity, and Zigbee for transmitting and receiving the information. Starting from wearable technologies — for example, Fit bits and Google Glass — to automobiles, companies are integrating Bluetooth capability into modern gadgets at a more noteworthy rate. Bluetooth's emerging omnipresence presents unique associations for the future. Bluetooth Low Energy can be used for the IoT because it is a suitable technology for transmitting and receiving the data using low energy wirelessly. However, there are some attacks described which also are of concern so as to possess added number of satisfied customers. Few attacks like Man in the middle (MITM), channel jamming, denial of service (DoS) and battery exhaustion along with some possible countermeasures described by various researchers are delineated in this overview.

The ability to develop innovative Business Models (BM) has become the cornerstone for the competitiveness of many successful business entities. Business Model Innovation (BMI) appears even more important as digitalization affect all areas of life and lead to new ways of working, communicating and cooperating in this digital world of Industry 4.0. Thanks to new technologies today businesses are connected in physical, digital and virtual networks, therefor the cooperation between businesses and academia and the use of competences across different domains will strengthen the BMI. One main challenge of digitization of BMI is modelling human behavior and cognitive processes into logical expressions that can be digitized and automated. The objective of this paper is to develop a conceptual model of an agile business model innovation learning environment which can be easily adapted by businesses utilizing sensors and machine learning techniques to observe, analyse and predict human behavior and facilitate the BMI development. The proposed technology backbone is a cloud-based sensing business model and sensing business model innovation room embedded with advanced mobile and wireless sensors, both environmental and wearable by the participants.

In this paper, we propose a novel feature extraction algorithm based on curvature analysis over the 3D data and the grayscale information extracted from the images. The performance of the method is tested on 3D mesh data, provided by Mars Orbiter Laser Altimeter (MOLA) and compared to benchmark research work. The experimental results demonstrate that the proposed method can achieve better accuracy in comparison with other crater detection methods and have smaller computational complexity.

With increased work load and unsuitable work shifts to survive in the fast paced world of today, people tend to lose sleep. Irregular sleep patterns and lack of sleep leads to drowsiness and fatigue. Drowsiness is perilous for the driver himself and for other drivers on the road and must be avoided for example by noise alerts in the car. This paper describes a method to detect drowsiness after implementing eye-tracking and mouth shape tracking in real-time. Viola-Jones algorithm is used to detect facial features in real-time. The proposed approach uses the detected facial features (i.e. eyes and mouth) based on Supervised Descent Method to find the blinking rate of a driver as well as for yawning detection. A decision, whether the driver is vigilant or not is then provided. Real time experiments based on publicly available dataset prove that the proposed method is highly efficient in finding the drowsiness and alerting the driver.

Automated expression recognition is a contemporary research field estimating human expressions from image or video data using computer algorithms combined with machine learning. This work proposes an algorithm for expression recognition including a feature extraction algorithm, consisting of log- Gabor filters followed by a feature selection based on sparse approximation of graph embedding. The classification is done on the selected features and is implemented using the Support Vector Machines classifier with radial basis kernel function. The algorithm is tested on the posed facial expressions image database Cohn-Kanade and provides competitive results compared to the state of the art.

The increasing role of spoken language interfaces in human-computer interaction applications has created conditions to facilitate a new area of research — namely recognizing the emotional state of the speaker through speech signals. This paper proposes a text independent method for emotion classification of speech signals used for the recognition of the emotional state of the speaker. Different feature selection criteria are explored and analyzed, namely Mutual Information Maximization (MIM) feature scoring criterion and its derivatives, to measure how potentially useful a feature or feature subset may be when used in a classifier. The proposed method employs different groups of low-level features, such as energy, zero-crossing rate, frequency bands in Mel scale, fundamental frequency or pitch, the delta- and delta-delta regression and statistical functions such as regression coefficients, extremums, moments etc., to represent the speech signals and a Neural Network classifier for the classification task. For the experiments the EMO-DB dataset is used with seven primary emotions including neutral. Results show that the proposed system yields an average accuracy of over 85% for recognizing 7 emotions with 5 of the best performing feature selection algorithms.

In this paper we present a comparative analysis of two algorithms for image representation with application to recognition of 3D face scans with the presence of facial expressions. We begin with processing of the input point cloud based on curvature analysis and range image representation to achieve a unique representation of the face features. Then, subspace projection using Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) is performed. Finally classification with different classifiers will be performed over the 3D face scans dataset with 61 subject with 7 scans per subject (427 scans), namely two "frontal", one "look-up", one "look-down", one "smile", one "laugh", one "random expression". The experimental results show a high recognition rate for the chosen database. They demonstrate the effectiveness of the proposed 3D image representations and subspace projection for 3D face recognition.

Solutions for caring for the elderly both efficacious and cost-effective are given by Ambient Assisted Living (AAL) systems that combine the research fields of intelligent systems and communication technologies. These systems are promising for the improvement of the quality of life of elderly and disabled people. One important characteristic of health and well-being is sleep. While sleep quantity is directly measurable, its quality has traditionally been assessed with subjective methods such as questionnaires. In this paper, we propose a non-intrusive sleep analyzer for real time detection of sleep anomalies, part of an effective AAL system. The proposed solution is based on combination of non-invasive sensors and an algorithm for sleep analysis with two stages - low and high level reasoning. It also offers the opportunity to include third party devices. Using the analyzer we can monitor basic sleep behavior and to detect sleep anomalies, which can serve as an important indicator for both mental and physical health.

In recent years, the EEG-based brain-computer interface (BCI) has become one of the most promising areas of research in computer science and robotics. Many internationally renewned research teams combining engineers and doctors, experts in neuroscience are trying to develop useful applications and devices offering disabled people to lead a normal life. Useful BCIs for disabled people suffering from Cerebral palsy, Parkinson's disease, Brain injury, Spinal cord injuries, Multiple sclerosis, Stroke, Post-polio syndrome should allow them to use all their existing brain and muscle abilities as control possibilities. In this paper we present a framework based on the mutimodal fusion approach of the user's electromyographic (EMG) and electroencephalographic (EEG) activities in a so called “Hybrid-BCI” (hBCI). Although EEG BCI alone yields good performance as already proved in many research papers, it is outperformed by the fusion of EEG and EMG. We investigate the influence of muscular fatigue on the EMG performance. Such a framework will allow a more reliable control and adaptation of the hBCI if the user get exhausted and loses concentration during the rehabilitation process. We focus our research in aims of improving the lives of many upper limb disabled individuals through a combination of current BCI technologies with existing assistive medical systems.