Deep Affect Prediction in-the-Wild: Aff-Wild Database and Challenge,Deep Architectures,and Beyond

International Journal of Computer Vision - Automatic understanding of human affect using visual signals is of great importance in everyday human–machine interactions. Appraising human...     

Optimum Power Allocation Based on Channel Conditions in Optical Satellite Downlinks

The integration of optical satellite links in the next-generation networks and in the fifth generation cellular systems has been proposed in order to guarantee the handling of the extreme data traffic growth and the high-pitched demand for networks’ resources. The optical satellite communication downlink is studied and more specifically, a geostationary satellite with multiple transmitters and an optical ground station with multiple receiving terminals are considered. In this contribution a novel power allocation methodology is proposed for the downlink. The allocation methodology takes into account the scintillation effects due to atmospheric turbulence and maximizes the ergodic network capacity under total expected power and peak power constraints. The analytical optimizing schemes are based on convex optimization methods and have been inspired by waterfilling algorithm. We present emulated power allocation results using real experimental downlink data from ARTEMIS optical satellite campaign and then we investigate the performance of the proposed algorithm with extended numerical results and comparison with other allocation policies. In particular, the new power allocation strategy achieves the highest spectral efficiency, according to the power constraints, for various channel conditions and attenuation profiles and also surpasses two simple baseband allocation methods by intelligently taking advantage of the number of channels and the total expected power.

     

A review of industrial wireless networks in the context of Industry 4.0

There have been many recent advances in wireless communication technologies, particularly in the area of wireless sensor networks, which have undergone rapid development and been successfully applied in the consumer electronics market. Therefore, wireless networks (WNs) have been attracting more attention from academic communities and other domains. From an industrial perspective, WNs present many advantages including flexibility, low cost, easy deployment and so on. Therefore, WNs can play a vital role in the Industry 4.0 framework, and can be used for smart factories and intelligent manufacturing systems. In this paper, we present an overview of industrial WNs (IWNs), discuss IWN features and related techniques, and then provide a new architecture based on quality of service and quality of data for IWNs. We also propose some applications for IWNs and IWN standards. Then, we will use a case from our previous achievements to explain how to design an IWN under Industry 4.0. Finally, we highlight some of the design challenges and open issues that still need to be addressed to make IWNs truly ubiquitous for a wide range of applications.     

VOC and carbonyl emissions from carpets: a comparative study using four types of environmental chambers

Emissions of Volatile Organic Compounds (VOCs) and carbonyls from carpets of different type (wool, synthetic) over a time period of three days at 23 degrees C, at 45% relative humidity, 0.5 air change rate and a loading factor of 0.4 m(2) m(-3) were measured. The experiments were carried out at four different environmental chambers (volumes of 0.02/0.28/0.45/30 m(3)). For TVOCs, maximum concentrations up to 2300 microg m(-3) (for carpet with synthetic backing) were found. Aromatic compounds e.g. benzene, toluene, the xylenes and styrene are emitted in relatively low concentrations, while for 4-phenylcyclohexene and 2, 2-butoxyethoxy-ethanol concentrations up to 170 and 320 microg m(-3), respectively, were measured. In all experiments, emission rates reached the maximum value within few hours from the beginning of the experiment. The emission rates of TVOCs from the same type of carpet measured with various types of chambers (0.02/0.28/0.45/30 m(3)), exhibited differences of up to 75%. Chamber concentrations of carbonyls (formaldehyde, acetaldehyde, acetone and propanale) are of a few microg m(-3). Acetone and formaldehyde reach concentrations up to 15 and 10 microg m(-3), respectively.     

Release of Zn, Ni, Cu, SO4(2-) and CrO4(2-) as a function of pH from cement-based stabilized/solidified refinery oily sludge and ash from incineration of oily sludge

A framework for the evaluation of leaching behavior of inorganic constituents from stabilized/solidified refinery oily sludge and ash produced from incineration of oily sludge with cement was employed. Metal and anion release as a function of pH was investigated. The leaching test consisted of multiple parallel extractions at pH range from 2 to 12. Remarkably good immobilization >98% was observed for metals of solidified ash at pH>6 and >93% of solidified oily sludge at pH>7. Sulfate leaching was high at pH range 2-12. The leaching behavior of metals and anions was simulated by VMINTEQ. The calculations showed that leaching behavior of Zn, Ni and Cu was controlled by chemical equilibrium and surface complexation onto ferrihydrite, at the pH range 2-12. The dominant solid phases that controlled metal leachability were metal hydroxides. The dominant mechanism that described sulfate leaching was found to be chemical equilibrium. Sulfate and also chromate leachability was controlled by Ettringite and Cr(VI)Ettringite as the major minerals affecting their release.     

Cement-based stabilization/solidification of oil refinery sludge: Leaching behavior of alkanes and PAHs

Stabilization/solidification is a process widely applied for the immobilization of inorganic constituents of hazardous wastes, especially for metals. Cement is usually one of the most common binders for that purpose. However, limited results have been presented on immobilization of hydrocarbons in cement-based stabilized/solidified petroleum solid waste. In this study, real oil refinery sludge samples were stabilized and solidified with various additions of I42.5 and II42.5 cement (Portland and blended cement, respectively) and subject to leaching. The target analytes were total petroleum hydrocarbons, alkanes and 16 polycyclic aromatic hydrocarbons of the EPA priority pollutant list. The experiments showed that the waste was confined in the cement matrix by macroencapsulation. The rapture of the cement structure led to the increase of leachability for most of the hydrocarbons. Leaching of n-alkanes from II42.5 cement-solidified samples was lower than that from I42.5 solidified samples. Leaching of alkanes in the range of n-C(10) to n-C(27) was lower than that of long chain alkanes (>n-C(27)), regardless the amount of cement addition. Generally, increasing the cement content in the solidified waste samples, increased individual alkane leachability. This indicated that cement addition resulted in destabilization of the waste. Addition of I42.5 cement favored immobilization of anthracene, benzo[a]anthracene, benzo[b]fluoroanthene, benzo[k]fluoroanthene, benzo[a]pyrene and dibenzo[a,h]anthracene. However, addition of II42.5 favored 5 out of 16, i.e., naphthalene, anthracene, benzo[b]fluoroanthene, benzo[k]fluoroanthene and dibenzo[a,h]anthracene.     

The use of an oil-absorber-bioscrubber system during biodegradation of sequentially alternating loadings of 1,2-dichloroethane and fluorobenzene in a waste gas

This work seeks to improve the robustness of vapour phase bioscrubbing by applying an absorber prior to a bioscrubber during the dynamic treatment of sequentially alternating loads of inhibitory pollutants. Rhizobiales sp. strain F11 and Xanthobacter autotrophicus sp. GJ10, exhibiting specific degradation capabilities for fluorobenzene (FB) and 1,2-dichloroethane (DCE), respectively, were used as a compound-strain model system. The stability of a combined oil-absorber-bioscrubber (OAB) was compared to the stability of a bioscrubber only (BO) system, during sequentially alternating periods (duration 3-6 d) of FB and DCE in the gas feed. The OAB achieved >66% degradation of FB, while in the BO system the FB removal efficiency dropped to 0% upon restoring FB feed after a 3 d FB starvation period. Following 6 d of FB starvation the BO failed to recover within 10 d, while the OAB required only 2 d to recover. In contrast, during the DCE treatment periods the OAB system did not show any advantage over the BO system. Further investigation showed that the F^- (a main metabolic product from FB degradation) has a strong inhibitory effect on strain GJ10 even at concentrations below . In the OAB system the inhibitory effect persisted for longer periods due to the absorber, which continued to supply FB to the system, and caused deterioration in the DCE removal efficiency. The inhibition of the microbial culture was confirmed by fluorescence in situ hybridisation (FISH), which showed that the activity of cells was reduced when only FB was fed. The results of this study have shown that in the presence of an inhibitory metabolic product the OAB system may not effectively improve the biological treatment of waste gas during sequential alternations in the feed of inhibitory pollutants.     

Cloud-assisted body area networks: state-of-the-art and future challenges

Body area networks (BANs) are emerging as enabling technology for many human-centered application domains such as health-care, sport, fitness, wellness, ergonomics, emergency, safety, security, and sociality. A BAN, which basically consists of wireless wearable sensor nodes usually coordinated by a static or mobile device, is mainly exploited to monitor single assisted livings. Data generated by a BAN can be processed in real-time by the BAN coordinator and/or transmitted to a server-side for online/offline processing and long-term storing. A network of BANs worn by a community of people produces large amount of contextual data that require a scalable and efficient approach for elaboration and storage. Cloud computing can provide a flexible storage and processing infrastructure to perform both online and offline analysis of body sensor data streams. In this paper, we motivate the introduction of Cloud-assisted BANs along with the main challenges that need to be addressed for their development and management. The current state-of-the-art is overviewed and framed according to the main requirements for effective Cloud-assisted BAN architectures. Finally, relevant open research issues in terms of efficiency, scalability, security, interoperability, prototyping, dynamic deployment and management, are discussed.     

Unsupervised Segmentation of Stereoscopic Video Objects: Constrained Segmentation Fusion Versus Greedy Active Contours

In this paper, we present a novel memory access reduction scheme (MARS) for two-dimension fast cosine transform (2-D FCT). It targets programmable DSPs with high memory-access latency. It reduces the number of memory accesses by: 1) reducing the number of weighting factors and 2) combining butterflies in vector-radix 2-D FCT pruning diagram from two stages to one stage with an efficient structure. Hardware platform based on general purpose processor is used to verify the effectiveness of the proposed method for vector-radix 2-D FCT pruning implementation. Experimental results validate the benefits of the proposed method with reduced memory access, less clock cycle and fewer memory space compared with the conventional implementation.