Abstract: Acoustic sensors, if embedded in smart street lights, can help in capturing the activities (car honking, sirens, events, traffic, etc.) in cities. Needless to say, the acoustic data from such scenarios are complex due to multiple audio streams originating from different events, and when decomposed to independent signals, the amount of retrieved data volume is small in quantity which is inadequate to train deep neural networks. So, in this paper, we address the two challenges: a) separating the mixed signals, and b) developing an efficient acoustic classifier under data paucity. So, to address these challenges, we propose an architecture with supervised deep learning, where the initial captured mixed acoustics data are analyzed with Fast Fourier Transformation (FFT), followed by filtering the noise from the signal, and then decomposed to independent signals by fast independent component analysis (Fast ICA). To address the challenge of data paucity, we propose a multi feature-based deep neural network with high performance that is reflected in our experiments when compared to the conventional convolutional neural network (CNN) and multi-layer perceptron (MLP).
Abstract: Background: Stroke is a disabling neurological disease. Rehabilitative therapies are important treatment methods. This clinical trial was done to compare the effects of virtual reality (VR) beside conventional rehabilitation versus conventional rehabilitation alone on the spasticity and motor function in stroke patients. Materials and methods: In this open-label randomized controlled clinical trial, 40 consecutive patients with stable first-ever ischemic stroke in the past three to 12 months that were referred to a rehabilitation clinic in Tehran, Iran in 2020 were enrolled. After signing the informed written consent form, subjects were randomly assigned by block randomization of five in each block as cases with 1:1 into two groups of 20 cases; conventional plus VR therapy group: 45-minute conventional therapy session plus 15-minute VR therapy, and conventional group: 60-minute conventional therapy session. VR rehabilitation is designed and developed with different stages. Outcomes were Modified Ashworth scale, Recovery Stage score for motor function, range of motion (ROM) of shoulder abduction/wrist extension, and patients’ satisfaction rate. Data were compared after study termination. Results: The satisfaction rate among the patients was significantly better in combination group (P = 0.003). Only wrist extension was varied between groups and was better in combination group. The variables generally had statistically significant difference (P < 0.05). Conclusion: VR plus conventional rehabilitation therapy is superior versus conventional rehabilitation alone on the wrist and elbow spasticity and motor function in patients with stroke.
Abstract: Various organisations often need to understand discussions on social media, such as what trending topics are and characteristics of the people engaged in the discussion. A number of approaches have been proposed to extract attributes that would characterise a discussion group. However, these approaches are largely based on supervised learning, and as such they require a large amount of labelled data. We propose an approach in this paper that does not require labelled data, but rely on lexical sources to detect meaningful attributes for online discussion groups. Our findings show an acceptable level of accuracy in detecting attributes for Twitter discussion groups.
Abstract: Despite the international commitment of Algeria towards the development of energy efficiency and renewable energy in the country, the internal energy demand has been continuously growing during the last decade due to the substantial increase of population and of living conditions, which in turn has led to an unprecedented expansion of the residential building sector. The RTB (Thermal Building Regulation) is the technical document that establishes the calculation framework for the thermal performance of buildings in Algeria, setting up minimum obligatory targets for the thermal performance of new buildings. An update of this regulation is due in the coming years and this paper discusses some proposals in this regard, with the aim to improve the energy efficiency of the building sector, particularly with regard to social housing. In particular, it proposes a methodology for drafting an energy performance label of new Algerian residential buildings, moving from the results of the thermal compliance verification and sizing of technical systems as defined in the RTB. Such an energy performance label – whose calculation method is briefly described in the paper – aims to raise citizens' awareness of the benefits of energy efficiency. It can represent the first step in a process of integrating technical installations into the calculation of the energy performance of buildings in Algeria.
Abstract: For successful re-integration, the individual offender must adapt and transform, which requires that the offender should adopt and internalise socially approved norms, attitudes, values, and beliefs. However, the offender’s labelling and community stigmatisation decide the destination of the offender. Community involvement in ex-offenders’ re-integration is an important issue in efforts to reduce recidivism and to control overcrowding in our correctional facilities. Crime is a social problem that requires society to come together to fight against it. This study was conducted in the Limpopo Province in Vhembe District Municipality within four local municipalities, namely Musina, Makhado, Mutale, and Thulamela. A total number of 30 participants were interviewed, and all were members of the Community Corrections Forums. This was necessitated by the fact that Musina is a very small area, which compelled the Department of Correctional Services to combine the two (Musina and Makhado) into one social re-integration entity. This is a qualitative research study where participants were selected through the use of purposive sampling. Participants were selected based on the value they would add to this study in order to achieve the objectives. The data collection method of this study was the focus group, which comprised of three groups of 10 participants each. Thulamela and Mutale local municipalities formed a group with (10) participants each, whereas Musina (2) and Makhado (8) formed another. Results indicate that the current situation is not conducive for re-integration to be successful. Participants raised many factors that need serious redress, namely offenders’ discrimination, lack of forgiveness by members of the community, which is fuelled by lack of community awareness due to the failure of the Department of Correctional Services in educating communities on ex-offenders’ re-integration.
Abstract: Estimating the 6D pose of objects is a core step for robot bin-picking tasks. The problem is that various objects are usually randomly stacked with heavy occlusion in real applications. In this work, we propose a method to regress 6D poses by predicting three points for each object in the 3D point cloud through deep learning. To solve the ambiguity of symmetric pose, we propose a labeling method to help the network converge better. Based on the predicted pose, an iterative method is employed for pose optimization. In real-world experiments, our method outperforms the classical approach in both precision and recall.
Abstract: Pseudomonas aeruginosa is an opportunistic pathogen that forms surface-associated microbial communities (biofilms) on artificial implant devices and on human tissue. Biofilm infections are difficult to treat with antibiotics, in part, because the bacteria in biofilms are physiologically heterogeneous. One measure of biological heterogeneity in a population of cells is to quantify the cellular concentrations of ribosomes, which can be probed with fluorescently labeled nucleic acids. The fluorescent signal intensity following fluorescence in situ hybridization (FISH) analysis correlates to the cellular level of ribosomes. The goals here are to provide computationally and statistically robust approaches to automatically quantify cellular heterogeneity in biofilms from a large library of epifluorescent microscopy FISH images. In this work, the initial steps were developed toward these goals by developing an automated biofilm detection approach for use with FISH images. The approach allows rapid identification of biofilm regions from FISH images that are counterstained with fluorescent dyes. This methodology provides advances over other computational methods, allowing subtraction of spurious signals and non-biological fluorescent substrata. This method will be a robust and user-friendly approach which will enable users to semi-automatically detect biofilm boundaries and extract intensity values from fluorescent images for quantitative analysis of biofilm heterogeneity.
Abstract: The present study investigates the use of the expression of refusal by native speakers of Jordanian Arabic (NSsJA) in different social situations (i.e. invitations, suggestions, and offers). It also investigates the influence of gender on the refusal realization patterns within the Jordanian culture to provide a better insight into the relation between situations, strategies and gender in the Jordanian culture. To that end, a group of 70 participants, including 35 male and 35 female students from different departments at the Hashemite University (HU) participated in this study using mixed methods (i.e. Discourse Completion Test (DCT), interviews and naturally occurring data). Data were analyzed in light of a developed coding scheme. The results showed that NSsJA preferred indirect strategies which mitigate the interaction such as "excuse, reason and, explanation" strategy more than other strategies which aggravate the interaction such as "face-threatening" strategy. Moreover, the analysis of this study has revealed a considerable impact of gender on the use of linguistic forms expressing refusal among NSsJA. Significant differences in the results of the Chi-square test relating the effect of participants' gender indicate that both males and females were conscious of the gender of their interlocutors. The findings provide worthwhile insights into the relation amongst types of communicative acts and the rapport between people in social interaction. They assert that refusal should not be labeled as face threatening act since it does not always pose a threat in some cases especially where refusal is expressed among friends, relatives and family members. They highlight some distinctive culture-specific features of the communicative acts of refusal.
Abstract: Sentiment analysis is a very active research topic.
Every day, Facebook, Twitter, Weibo, and other social media,
as well as significant e-commerce websites, generate a massive
amount of comments, which can be used to analyse peoples
opinions or emotions. The existing methods for sentiment analysis
are based mainly on sentiment dictionaries, machine learning, and
deep learning. The first two kinds of methods rely on heavily
sentiment dictionaries or large amounts of labelled data. The third
one overcomes these two problems. So, in this paper, we focus
on the third one. Specifically, we survey various sentiment analysis
methods based on convolutional neural network, recurrent neural
network, long short-term memory, deep neural network, deep belief
network, and memory network. We compare their futures, advantages,
and disadvantages. Also, we point out the main problems of
these methods, which may be worthy of careful studies in the
future. Finally, we also examine the application of deep learning in
multimodal sentiment analysis and aspect-level sentiment analysis.
Abstract: Noise estimation is essential in today wireless systems
for power control, adaptive modulation, interference suppression and
quality of service. Deep learning (DL) has already been applied in the
physical layer for modulation and signal classifications. Unacceptably
low accuracy of less than 50% is found to undermine traditional
application of DL classification for SNR prediction. In this paper,
we use divide-and-conquer algorithm and classifier fusion method
to simplify SNR classification and therefore enhances DL learning
and prediction. Specifically, multiple CNNs are used for classification
rather than a single CNN. Each CNN performs a binary classification
of a single SNR with two labels: less than, greater than or equal.
Together, multiple CNNs are combined to effectively classify over a
range of SNR values from −20 ≤ SNR ≤ 32 dB.We use pre-trained
CNNs to predict SNR over a wide range of joint channel parameters
including multiple Doppler shifts (0, 60, 120 Hz), power-delay
profiles, and signal-modulation types (QPSK,16QAM,64-QAM). The
approach achieves individual SNR prediction accuracy of 92%,
composite accuracy of 70% and prediction convergence one order
of magnitude faster than that of traditional estimation.
Abstract: This paper presents and benchmarks a number of
end-to-end Deep Learning based models for metaphor detection in
Greek. We combine Convolutional Neural Networks and Recurrent
Neural Networks with representation learning to bear on the metaphor
detection problem for the Greek language. The models presented
achieve exceptional accuracy scores, significantly improving the
previous state-of-the-art results, which had already achieved accuracy
0.82. Furthermore, no special preprocessing, feature engineering or
linguistic knowledge is used in this work. The methods presented
achieve accuracy of 0.92 and F-score 0.92 with Convolutional
Neural Networks (CNNs) and bidirectional Long Short Term Memory
networks (LSTMs). Comparable results of 0.91 accuracy and 0.91
F-score are also achieved with bidirectional Gated Recurrent Units
(GRUs) and Convolutional Recurrent Neural Nets (CRNNs). The
models are trained and evaluated only on the basis of training tuples,
the related sentences and their labels. The outcome is a state-of-the-art
collection of metaphor detection models, trained on limited labelled
resources, which can be extended to other languages and similar
tasks.
Abstract: Engagement is one of the most important factors in determining successful outcomes and deep learning in students. Existing approaches to detect student engagement involve periodic human observations that are subject to inter-rater reliability. Our solution uses real-time multimodal multisensor data labeled by objective performance outcomes to infer the engagement of students. The study involves four students with a combined diagnosis of cerebral palsy and a learning disability who took part in a 3-month trial over 59 sessions. Multimodal multisensor data were collected while they participated in a continuous performance test. Eye gaze, electroencephalogram, body pose, and interaction data were used to create a model of student engagement through objective labeling from the continuous performance test outcomes. In order to achieve this, a type of continuous performance test is introduced, the Seek-X type. Nine features were extracted including high-level handpicked compound features. Using leave-one-out cross-validation, a series of different machine learning approaches were evaluated. Overall, the random forest classification approach achieved the best classification results. Using random forest, 93.3% classification for engagement and 42.9% accuracy for disengagement were achieved. We compared these results to outcomes from different models: AdaBoost, decision tree, k-Nearest Neighbor, naïve Bayes, neural network, and support vector machine. We showed that using a multisensor approach achieved higher accuracy than using features from any reduced set of sensors. We found that using high-level handpicked features can improve the classification accuracy in every sensor mode. Our approach is robust to both sensor fallout and occlusions. The single most important sensor feature to the classification of engagement and distraction was shown to be eye gaze. It has been shown that we can accurately predict the level of engagement of students with learning disabilities in a real-time approach that is not subject to inter-rater reliability, human observation or reliant on a single mode of sensor input. This will help teachers design interventions for a heterogeneous group of students, where teachers cannot possibly attend to each of their individual needs. Our approach can be used to identify those with the greatest learning challenges so that all students are supported to reach their full potential.
Abstract: In this study, an automated building and tree detection
method is proposed using DSM data and true orthophoto image. A
multiscale matched filtering is used on DSM data. Therefore, first
watershed transform is applied. Then, Otsu’s thresholding method
is used as an adaptive threshold to segment each watershed region.
Detected objects are masked with NDVI to separate buildings and
trees. The proposed method is able to detect buildings and trees
without entering any elevation threshold. We tested our method on
ISPRS semantic labeling dataset and obtained promising results.
Abstract: Bone metastases are observed in a wide range of cancers leading to intolerable pain. While early detection can help the physicians in the decision of the type of treatment, various radiopharmaceuticals using phosphonates like 68Ga-EDTMP have been developed. In this work, due to the importance of absorbed dose, human absorbed dose of this new agent was calculated for the first time based on biodistribution data in Wild-type rats. 68Ga was obtained from 68Ge/68Ga generator with radionuclidic purity and radiochemical purity of higher than 99%. The radiolabeled complex was prepared in the optimized conditions. Radiochemical purity of the radiolabeled complex was checked by instant thin layer chromatography (ITLC) method using Whatman No. 2 paper and saline. The results indicated the radiochemical purity of higher than 99%. The radiolabelled complex was injected into the Wild-type rats and its biodistribution was studied up to 120 min. As expected, major accumulation was observed in the bone. Absorbed dose of each human organ was calculated based on biodistribution in the rats using RADAR method. Bone surface and bone marrow with 0.112 and 0.053 mSv/MBq, respectively, received the highest absorbed dose. According to these results, the radiolabeled complex is a suitable and safe option for PET bone imaging.
Abstract: The neural network quantization is highly desired
procedure to perform before running neural networks on mobile
devices. Quantization without fine-tuning leads to accuracy drop of
the model, whereas commonly used training with quantization is done
on the full set of the labeled data and therefore is both time- and
resource-consuming. Real life applications require simplification and
acceleration of quantization procedure that will maintain accuracy of
full-precision neural network, especially for modern mobile neural
network architectures like Mobilenet-v1, MobileNet-v2 and MNAS. Here we present a method to significantly optimize training with
quantization procedure by introducing the trained scale factors for
discretization thresholds that are separate for each filter. Using the
proposed technique, we quantize the modern mobile architectures of
neural networks with the set of train data of only ∼ 10% of the
total ImageNet 2012 sample. Such reduction of train dataset size and
small number of trainable parameters allow to fine-tune the network
for several hours while maintaining the high accuracy of quantized
model (accuracy drop was less than 0.5%). Ready-for-use models and
code are available in the GitHub repository.
Abstract: A significant discussion on the topic of blockchain as a solution to the issues of intellectual property highlights the relevance that this topic holds. Some experts label this technology as destructive since it holds immense potential to change course of traditional practices. The extent and areas to which this technology can be of use are still being researched. This paper provides an in-depth review on the intellectual property and blockchain technology. Further it explores what makes blockchain suitable for intellectual property, the practical solutions available and the support different governments are offering. This paper further studies the framework of universities in context of its outputs and how can they be streamlined using blockchain technology. The paper concludes by discussing some limitations and future research question.
Abstract: Bioactive glasses (BGs) are a group of surface-reactive biomaterials used in clinical applications as implants or filler materials in the human body to repair and replace diseased or damaged bone. Sol-gel technique was employed to prepare a SiO2-CaO-P2O5 glass with nominal composition of 58S BG with the addition of Sr and Li modifiers which imparts special properties to the BG. The effect of simultaneous addition of Sr and Li on bioactivity and biocompatibility, proliferation, alkaline phosphatase (ALP) activity of osteoblast cell line MC3T3-E1 and antibacterial property against methicillin-resistant Staphylococcus aureus (MRSA) bacteria were examined. BGs were characterized by X-ray diffraction, Fourier transform infrared spectroscopy, scanning electron microscopy before and after soaking the samples in the simulated body fluid (SBF) for different time intervals to characterize the formation of hydroxyapatite (HA) formed on the surface of BGs. Structural characterization indicated that the simultaneous presence of 5% Sr and 5% Li in 58S-BG composition not only did not retard HA formation because of opposite effect of Sr and Li of the dissolution of BG in the SBF but also, stimulated the differentiation and proliferation of MC3T3-E1s. Moreover, the presence of Sr and Li on dissolution of the ions resulted in an increase in the mean number of DAPI-labeled nuclei which was in good agreement with live/dead assay. The result of antibacterial tests revealed that Sr and Li-substituted 58S BG exhibited a potential antibacterial effect against MRSA bacteria. Because of optimal proliferation and ALP activity of MC3T3-E1cells, proper bioactivity and high antibacterial potential against MRSA, BG-5/5 is suggested as a multifunctional candidate for bone tissue engineering.
Abstract: When the public space is in an emergency, how to quickly establish spatial cognition and emergency shelter in the closed underground space is the urgent task. This study takes Taipei Station as the research base and aims to apply the use of Internet of things (IoT) application for underground evacuation mobility design. The first experiment identified passengers' evacuation behaviors and spatial cognition in underground spaces by wayfinding tasks and thinking aloud, then defined the design conditions of User Interface (UI) and proposed the UI design. The second experiment evaluated the UI design based on passengers' evacuation behaviors by wayfinding tasks and think aloud again as same as the first experiment. The first experiment found that the design conditions that the subjects were most concerned about were "map" and hoping to learn the relative position of themselves with other landmarks by the map and watch the overall route. "Position" needs to be accurately labeled to determine the location in underground space. Each step of the escape instructions should be presented clearly in "navigation bar." The "message bar" should be informed of the next or final target exit. In the second experiment with the UI design, we found that the "spatial map" distinguishing between walking and non-walking areas with shades of color is useful. The addition of 2.5D maps of the UI design increased the user's perception of space. Amending the color of the corner diagram in the "escape route" also reduces the confusion between the symbol and other diagrams. The larger volume of toilets and elevators can be a judgment of users' relative location in "Hardware facilities." Fire extinguisher icon should be highlighted. "Fire point tips" of the UI design indicated fire with a graphical fireball can convey precise information to the escaped person. "Fire point tips" of the UI design indicated fire with a graphical fireball can convey precise information to the escaped person. However, "Compass and return to present location" are less used in underground space.
Abstract: Brachistinae Föerster, 1862 is a subfamily of the family Braconidae (order Hymenoptera) with about 410 species distributed all around the world. Brachistinae includes the genera, Eubazus Nees von Esenbeck 1814, Foersteria Szépligeti 1896, Chelostes van Achterberg 1990, Triaspis Haliday 1835 and Schizoprymnus Förster 1862. Members of the subfamily live as parasitoids on the families Curculionidae and Apionidae (Coleoptera), which also include very important agricultural pests. In generally, members of the genus Triaspis are poorly known biologically. The genus is represented by 37 species in the West Palearctic region and 118 species worldwide. Adult specimens of Triaspis were collected from as wide a range of habitats as possible at different altitudes in different parts of Turkey between 1982 and 2010. Samples collected from short plants using standard insect sweeping nets were transferred into tubes containing 70% ethanol and labelled following their preparations according to museum techniques. Seven Triaspis species have been reported from Turkey in this study. Five of these species are new to the fauna of Turkey.
Abstract: Satellite imagery is one of the emerging technologies which are extensively utilized in various applications such as detection/extraction of man-made structures, monitoring of sensitive areas, creating graphic maps etc. The main approach here is the automated detection of buildings from very high resolution (VHR) optical satellite images. Initially, the shadow, the building and the non-building regions (roads, vegetation etc.) are investigated wherein building extraction is mainly focused. Once all the landscape is collected a trimming process is done so as to eliminate the landscapes that may occur due to non-building objects. Finally the label method is used to extract the building regions. The label method may be altered for efficient building extraction. The images used for the analysis are the ones which are extracted from the sensors having resolution less than 1 meter (VHR). This method provides an efficient way to produce good results. The additional overhead of mid processing is eliminated without compromising the quality of the output to ease the processing steps required and time consumed.