Search

With the progress of human society, the consumption of energy materials is increasing in particularly the fossil fuels, which accounts for 84.7% of the total global consumption in 2018. Thus, excessive exploitation of fossil fuels not only brings about an increasingly serious energy crisis, but also leads to a series of environmental problems, such as the air quality deterioration and soil acidification, which are not conducive to the sustainable development of society. Since the 21st Annual Conference of the China Association for Science and Technology in 2019, “renewable synthetic fuels” has become one of the focuses of attention, and it also puts forward higher requirements for the research and development of renewable energy. About 60 years ago, many plants, insects, microorganisms and microalgae were found to metabolize hydrocarbons. In 2010, Schirmer’s groups recognized the alkane biosynthesis pathways of marine chlorobacteria, which aroused great interest in hydrocarbon biosynthesis. Recently, more proteases to catalyze the synthesis of hydrocarbons have been successively discovered, such as aldehyde deformylase (ADO, CYP4G and CER), fatty acid decarboxylase (UndA, UndB and OleT), polyketone synthase (Ols) and thiolytic enzyme complex system (OleABCD). In the past decades, many studies have been conducted to explore the biosynthesis of hydrocarbon, and useful information regarding the reaction mechanism has been obtained, including the kinetic rate constants and the crystal structures of the enzymes. In addition, on the basis of crystal structure, quantum chemistry method (QM) and molecular dynamics simulations (MD) were also used to explore the reaction mechanism of fatty acid decarboxylation. Since most enzymes involved in the hydrocarbon biosynthesis are iron-containing enzymes, which employ the Fe-coordinate dioxygen to trigger the reaction. In general, the whole reaction cycle contains the binding of substrates and dioxygen, and decarboxylation usually involves the hydrogen atom abstraction and breaking of C−C bond. During the reaction, complex electron transfer takes place among the Fe center, O2 and substrate, and many factors were found to affect the catalytic activity. According to the previous experimental and theoretical studies, different enzymes show different activity and substrate scope. In general, the catalytic efficiency is still very low for most enzymes, and thus, understanding the mechanism of these catalytic reactions is undoubtedly an important basis for engineering optimization and all applications, and can also provide ideas for the design and synthesis of biomimetic catalysts. In this paper, we review the research progress of five representative iron enzymes that catalyze the biosynthesis of aliphatic hydrocarbon, including OleT, UndA, UndB, ADO and CYP4G, and discuss the future development and application of enzyme catalysis in renewable synthetic fuels. This paper may offer some ideas and guidelines for sustainable energy supply in the future.

Background Alzheimer's disease (AD) is one of the most prevalent neurodegenerative disorders. Enhancing hippocampal neurogenesis by promoting proliferation and differentiation of neural stem cells (NSCs) is a promising therapeutic strategy for AD. 20(S)-protopanaxadiol (PPD) and oleanolic acid (OA) are small, bioactive compounds found in ginseng that can promote NSC proliferation and neural differentiation in vitro. However, it is currently unknown whether PPD or OA can attenuate cognitive deficits by enhancing hippocampal neurogenesis in vivo in a transgenic APP/PS1 AD mouse model. Here, we administered PPD or OA to APP/PS1 mice and monitored the effects on cognition and hippocampal neurogenesis. Methods We used the Morris water maze, Y maze, and open field tests to compare the cognitive capacities of treated and untreated APP/PS1 mice. We investigated hippocampal neurogenesis using Nissl staining and BrdU/NeuN double labeling. NSC proliferation was quantified by Sox2 labeling of the hippocampal dentate gyrus. We used western blotting to determine the effects of PPD and OA on Wnt/GSK3β/β-catenin pathway activation in the hippocampus. Results Both PPD and OA significantly ameliorated the cognitive impairments observed in untreated APP/PS1 mice. Furthermore, PPD and OA significantly promoted hippocampal neurogenesis and NSC proliferation. At the mechanistic level, PPD and OA treatments resulted in Wnt/GSK-3β/β-catenin pathway activation in the hippocampus. Conclusion PPD and OA ameliorate cognitive deficits in APP/PS1 mice by enhancing hippocampal neurogenesis, achieved by stimulating the Wnt/GSK-3β/β-catenin pathway. As such, PPD and OA are promising novel therapeutic agents for the treatment of AD and other neurodegenerative diseases.

Global warming and depletion of fossil fuels have urged the world to control their impacts and find clean and green energy sources. Due to this paradigm shift, renewable energy resources (RESs) have received tremendous attention around the globe and amongst other RESs, wind energy has been proving itself a promising renewable resource. However, still, the proper selection of wind turbines (WTs) and their installation is considered a major challenge. Presently, due to the economic boom of Pakistan as a regional power under the China Pakistan Economic Corridor (CPEC) and increasing energy demand, it needs to meet the energy shortage. This paper aims to conduct a pre-feasibility ainvestigation of wind farm exploitation at Gwadar and Quetta, major cities of Baluchistan province in Pakistan by harnessing the wind energy potential and electricity generation cost through different WT models. The wind speed (WS) data is recorded by a professional wind mast at different measurement heights during F.Y 2016 and analyzed using Five approaches, Modified Maximum Likelihood Method (MMLM), Graphical Method (GM), Empirical Method of Lysen (EML), Empirical Method of Jestus (EMJ) and Energy Pattern Factor Method (EPF) at the proposed site and EPF shows the best performance than other approaches. Different WTs models are also evaluated and out of them, the GW121 WT gives the lowest cost in dollars and lowest payback period at each site so it can be recommended as one of the most feasible WT for electricity generation for powering local communities at the proposed sites.

Facial expression recognition from video sequences is currently an interesting research topic in computer vision, pattern recognition, artificial intelligence, etc. Considering the problem of seman...

Introduction Type 2 diabetes mellitus erectile dysfunction (T2DMED) is one of the common complications of type 2 diabetes mellitus (T2DM). Icariside II (ICA II), a flavonoid derived from Epimedium, has been shown to improve erectile function in T2DMED rats. Aim To investigate the effect of ICA II and metformin (MET) on penile erectile function, mitochondrial autophagy, glucose-lipid metabolism in rats with T2DMED. Methods In the control and T2DMED groups, rats were administered normal saline. In the MET group, rats were administered MET for 0.2 g/kg/day. In the ICA II+MET group, rats were administered ICA II for 10 mg/kg/day and MET for 0.2 g/kg/day. Results The number of mating rats, number of erectile rats, erection rate, erection frequency, intracorneal pressure, and intracorneal pressure/mean arterial pressure in the ICA II+MET group and control group were significantly higher than corresponding values in than T2DMED group. The absolute values of fasting plasma glucose, glycated haemoglobin in the ICA II+MET group, MET group, and control group were significantly lower than in the T2DMED group. The advanced glycation end product (AGE) values in the ICA II+MET group and the MET group were lower than in the T2DMED group. The receptors for the AGE values and angiotensin II values in the ICA II+MET group were lower than in the T2DMED and MET groups. The high-density lipoprotein values, testosterone values, nitric oxide synthase activity, and cyclic guanosine monophosphate content in the ICA II+MET and control groups were higher than in the T2DMED group. The low-density lipoprotein values, triglyceride values, estradiol values, and total cholesterol values in the ICA II+MET and control groups were lower than in the T2DMED group. Conclusion ICA II could increase erectile function and smooth muscle cell/collagen fibril proportions, decreased mitochondrial autophagy, and AGE concentrations and improve lipid metabolism, nitric oxide synthase activity, cyclic guanosine monophosphate content, testosterone, estradiol, and Ang II in rat with T2DMED.

Automatic speech emotion recognition (SER) is a challenging component of human-computer interaction (HCI). Existing literatures mainly focus on evaluating the SER performance by means of training and testing on a single corpus with a single language setting. However, in many practical applications, there are great differences between the training corpus and testing corpus. Due to the diversity of different speech emotional corpus or languages, most previous SER methods do not perform well when applied in real-world cross-corpus or cross-language scenarios. Inspired by the powerful feature learning ability of recently-emerged deep learning techniques, various advanced deep learning models have increasingly been adopted for cross-corpus SER. This paper aims to provide an up-to-date and comprehensive survey of cross-corpus SER, especially for various deep learning techniques associated with supervised, unsupervised and semi-supervised learning in this area. In addition, this paper also highlights different challenges and opportunities on cross-corpus SER tasks, and points out its future trends.

We, the Editors and Publisher of the journal Artificial Cells, Nanomedicine, and Biotechnology, have retracted the following article: Genggang Cui, Mingming Shao, Xingzhou Gu, Hongbo Guo, Shiqing Zhang, Jianlei Lu & Hongbin Ma (2019) The value of FGF9 as a novel biomarker in the diagnosis of prostate cancer. Artificial Cells, Nanomedicine, and Biotechnology, 47:1, 2241–2245, DOI: 10.1080/21691401.2019.1620250 It has come to our attention that the full authorship list and affiliations for this manuscript, including the study site and ethics committee, were changed after the article was submitted. We have contacted the authors for an explanation, but we have not received a response within the requested timeframe. As determining authorship and the location of where the research was conducted is core to the integrity of published work, we are therefore retracting the article. The authors listed in this publication have been sent notification. We have been informed in our decision-making by our policy on publishing ethics and integrity and the COPE guidelines on retractions. The retracted article will remain online to maintain the scholarly record, but it will be digitally watermarked on each page as ‘Retracted’.

One key challenging issues of facial expression recognition (FER) in video sequences is to extract discriminative spatiotemporal video features from facial expression images in video sequences. In this paper, we propose a new method of FER in video sequences via a hybrid deep learning model. The proposed method first employs two individual deep convolutional neural networks (CNNs), including a spatial CNN processing static facial images and a temporal CN network processing optical flow images, to separately learn high-level spatial and temporal features on the divided video segments. These two CNNs are fine-tuned on target video facial expression datasets from a pre-trained CNN model. Then, the obtained segment-level spatial and temporal features are integrated into a deep fusion network built with a deep belief network (DBN) model. This deep fusion network is used to jointly learn discriminative spatiotemporal features. Finally, an average pooling is performed on the learned DBN segment-level features in a video sequence, to produce a fixed-length global video feature representation. Based on the global video feature representations, a linear support vector machine (SVM) is employed for facial expression classification tasks. The extensive experiments on three public video-based facial expression datasets, i.e., BAUM-1s, RML, and MMI, show the effectiveness of our proposed method, outperforming the state-of-the-arts.

An-Gong-Niu-Huang Wan (AGNHW), a famous formula in traditional Chinese medicine, has been clinically used for centuries for treating cerebral diseases, but the protective effects of pre-treatment with AGNHW on cerebral ischemia have not yet been reported. The present study aimed to test such protective effects and elucidate the underlying mechanisms on cerebral ischemia in rats by phenotypic approaches (i.e. including the neurological functional score, cerebral infarct area, neuron apoptosis, and brain oxidative stress status) and target-based approaches (i.e. involving the GSK-3β/HO-1 pathway). AGNHW was administered orally at the doses of 386.26, 772.52, and 1545.04 mg/kg respectively for 7 days to male Sprague-Dawley rats and then cerebral ischemia was induced by middle cerebral artery occlusion (MCAO) for 1.5 h. Pre-treatment with AGNHW significantly ameliorated ischemic damage to the brain in a dose-dependent manner, including reduction of the neurological deficit score and infarct area. AGNHW pre-treatment increased the number of Nissl+ cells, NeuN+ and DCX+ cells, and decreased the number of Tunel+ cells. Moreover, AGNHW reversed the up-regulation of ROS and MDA induced by cerebral ischemia. AGNHW pre-treatment increased the expression of p-GSK-3β(Ser9)/GSK-3β (glycogen synthase kinase-3β) ratio and heme oxygenase-1 (HO-1). These results firstly revealed that short-term pre-treatment of AGNHW could significantly protect the rats from injury caused by cerebral ischemia-reperfusion, which support further clinical studies for disease prevention. The in vivo protective effect of AGNWH pre-treatment could be associated with its antioxidant properties by the activation of GSK-3β-mediated HO-1 pathway.

COVID‐19 is spreading rapidly worldwide, posing great threats to public health and economy. This study aims to examine how the transmission of COVID‐19 is modulated by climate conditions, which is of great importance for better understanding of the seasonal feature of COVID‐19. Constrained by the accurate observations we can make, the basic reproduction numbers (R 0) for each country were inferred and linked to temperature and relative humidity (RH) with statistical analysis. Using R 0 as the measure of COVID‐19 transmission potential, we find stronger transmission of COVID‐19 under mildly warm (0°C < T 60%) climate conditions, while extremely low (T 20°C) temperature or a dry climate (RH, The distribution of R 0 with temperature (left) and relative humidity (right).

The outcomes of thoracic surgery for patients with stage IV non-small-cell lung cancer (NSCLC) are controversial and uncertain.The National Cancer Institute's Surveillance, Epidemiology, and End Results was queried for patients with stage IV NSCLC, including those treated with surgery-participated therapy modalities. Overall survival (OS) was evaluated using a variety of statistical analyses.The analysis was carried out for 90,982 patients from 1975 to 2016 who had been diagnosed as stage IV NSCLC. Propensity score-matched (PSM) analyses that were well-balanced with all the important confounding covariates revealed improved OS (median survival time [MST]) with patients receiving surgery versus non-surgery (MST: 15 versus 8 months, P 0.001); undergoing surgery plus chemotherapy versus chemotherapy (MST: 19 versus 11 months, P 0.001); and having surgery plus chemoradiation versus chemoradiation (MST: 18 versus 11 months, P 0.001). Sequential landmark analyses for long-term survivors of ≥1 and ≥3 years all indicated improved OS (P 0.001) on univariate and multivariate analyses for the patients receiving the three surgery-related treatment patterns listed earlier, relative to the corresponding surgery-absent treatment modalities. For synchronous presentations of varied treatment paradigms, surgical intervention significantly led to increased OS (MST, months) benefits following treatment paradigms: surgery plus chemotherapy (22), surgery plus chemoradiation (18), chemotherapy (10), surgery only (9), chemoradiation (9), surgery plus radiation (6) and radiation alone (2). The subgroup analysis demonstrated that the elevated OS associated with local thoracic surgery in addition to chemotherapy (versus chemotherapy) or chemoradiation (versus chemoradiation) fell in the subcategories of T0-3, N0-2 and 0-1 (metastatic sites) tumours. The comparison of the aforementioned two types of treatment patterns indicated that the optimal patients for the surgery were those with any combination of T1-4, N0-3, Msite0-1 and adeno- or squamous carcinoma.The patients with T1-4, N0-3, Msite0-1 and adeno- or squamous carcinoma of stage IV NSCLC had a longer OS with local thoracic surgery in combination with chemotherapy or chemoradiation.

Fibrosis is unregulated tissue repair in damaged or diseased organs, and the accumulation of excess extracellular matrix (ECM) impacts the structure and functions of organs, leading to death. Fibrosis is usually triggered by inflammation and tissue damage, and inflammatory mediators stimulate the proliferation of myofibroblasts and the excessive production of ECM. The IL-10 family cytokines play important roles in the development of fibrosis, and its member IL-22 has recently attracted specific attention. IL-22 plays great roles in preventing pathogens invasion and tissue damage, as well as making a contribution to pathogenic processes. Increasing evidence suggested that IL-22 is a key molecule in tissue repair, proliferation and mucosal barrier defense, and it has also been suggested to play both pro-fibrotic and anti-fibrotic roles in tissues. In this review, we summarized the pro-fibrotic and anti-fibrotic functions of IL-22 in various organs which may be of great significance for the development of potential therapeutic strategies for fibrosis-related diseases.

Background A steady progress on schistosomiasis control in the Peoples’ Republic of China (P.R. China) was achieved and broadened into the twelve-year medium and long term national plan (MLNP) which marled the implementation of an integrated control strategy across all endemic areas in P.R. China in 2004. To understand the endemic trends of schistosomiasis to assess the effectiveness of an integrated strategy, we conducted an analysis of schistosomiasis surveillance data spanned from 2005 to 2015. Methods The schistosomiasis sentinel surveillance data from sentinel sites were collected and analyzed from 2005 to 2015. In these sentinel sites, residents aged 6 years or above were screened annually by indirect hemagglutination assay (IHA), while only antibody positives were followed by stool examination either Kato-katz method (KK) and/or hatching technique (HT). Domestic animals raised in sentinel sites were examined by HT for confirming the infection of schistosomes. Snail investigation was conducted each year through systematic sampling method combined with environmental sampling method. The snails collected from field were tested by microscopic dissection method. The infection rates of schistosomes in residents, domestic animals and snails, as well as the indicators reflecting the snails’ distribution were calculated and analyzed. ANOVA analysis was used to examine the changes of the number of eggs per gram feces in population and Chi-square test was used to examine any change in proportions among groups. Results A total of 148 902 residents from sentinel sites attended this study and 631 676 blood samples were examined by IHA test during the 11 covered years. The annual average antibody positive rates presented a significant decrease trends, from 17.48% (95% CI: 17.20–17.75%) in 2005 to 5.93% (95% CI: 5.71–6.15%) (χ2 = 8890.47, P CI: 1.96–2.17%) to 0.13% (95% CI: 0.09–0.16%), accompanied by significant decrease of infection intensity in population. In 2015, the stool positives were only found in farmers, fishermen and boatmen with infection rate of 0.16% (95% CI: 0.11–0.20%), 0.17% (95% CI: 0–0.50%) respectively. The infection rate of schistosomes in domestic animals dropped from 9.42% (538/5711, 95% CI: 8.66–10.18%) to 0.08% (2/2360, 95% CI: 0–0.20%) from 2005 to 2015. Infections were found in eight species of domestic animals at the beginning of surveillance while only two cattle were infected in 2015. Totally 98 ha of new snail habitats were found, while 94.90% (93/98) distributed in lake and marshland regions. The percentage of frames with snails decreased from 16.96% (56 884/33 5391, 95% CI: 16.83–17.09%) in 2005 to 4.28% (18 121/423 755, 95% CI: 4.22–4.34%) in 2014, with a slightly increase in 2015. Meanwhile, the infection rate of schistosomes in snails was decreased from 0.26% (663/256 531, 95% CI: 0.24–0.28%) to zero during 2005–2015. Conclusions The infection rate of schistosomes declined significantly, providing evidence that the goal of the MLNP was achieved. Elimination of schistosomiasis as a public health problem defined as WHO was also reached in P.R. China nationwide. Surveillance-response system should be improved and strengthened to realize the final goal of schistosomiasis elimination.

Background Icariside II (ICAII) is a flavonoid isolated from herb Epimedium that has been shown to improve erectile function in rats. However, ICAII's underlying mechanism remains unclear. Methods Type 2 diabetes mellitus erectile dysfunction (T2DMED) rats were induced by single intraperitoneal injection of 25 mg/kg streptozotocin (STZ) and fed a high-fat, high-sugar, and high-calorie diet for 8 weeks. In the control and T2DMED groups, rats were administered with normal saline; in the metformin (MET) group, rats were administered with MET at 0.2 g/kg/day; and in the ICAII + MET group, rats were administered with ICA II at 10 mg/kg/day and MET for 0.2 g/kg/day. The deposition of advanced glycation end products (AGEs), expression of receptor for AGEs (RAGEs), reactive oxygen species (ROS), and superoxide dismutase (SOD) activity, and corpus cavernosum smooth muscle cells (CCSMCs) mitochondrial autophagy were measured. We also evaluated the expression of LC3-II/I, Beclin-1, P70S6K, PI3K, AKT, mTOR, phospho-AKT, phospho-mTOR, and phospho-P70S6K. Results ICAII and MET can improve erectile function, and decrease the levels of fasting plasma glucose (FPG), hemoglobin A1c (HbA1c), and AGEs in rats with T2DMED. Furthermore, ICAII and MET can decrease excessive CCSMC mitochondrial autophagy and the level of RAGE and oxidant stress in rats with T2DMED. ICAII and MET may enhance signaling via the PI3K-AKT-mTOR pathway in order to reduce the excessive mitochondrial autophagy of CCSMCs. Conclusions ICAII may effectively improve penile erectile function via decreasing excessive CCSMCs mitochondrial autophagy, RAGE, and oxidant stress. Furthermore, ICAII may enhance signaling via the PI3K-AKT-mTOR pathway in order to reduce excessive CCSMC mitochondrial autophagy.

In order to save people’s shopping time and reduce labor cost of supermarket operations, this paper proposes to design a supermarket service robot based on deep convolutional neural networks (DCNNs). Firstly, according to the shopping environment and needs of supermarket, the hardware and software structure of supermarket service robot is designed. The robot uses a robot operating system (ROS) middleware on Raspberry PI as a control kernel to implement wireless communication with customers and staff. So as to move flexibly, the omnidirectional wheels symmetrically installed under the robot chassis are adopted for tracking. The robot uses an infrared detection module to detect whether there are commodities in the warehouse or shelves or not, thereby grasping and placing commodities accurately. Secondly, the recently-developed single shot multibox detector (SSD), as a typical DCNN model, is employed to detect and identify objects. Finally, in order to verify robot performance, a supermarket environment is designed to simulate real-world scenario for experiments. Experimental results show that the designed supermarket service robot can automatically complete the procurement and replenishment of commodities well and present promising performance on commodity detection and recognition tasks.

Sepsis is a leading cause of mortality in intensive care units and costs hospitals billions of dollars annually worldwide. Predicting survival time for sepsis patients is a time-critical prediction problem. Considering the useful sequential information for sepsis development, this paper proposes a time-critical topic model (TiCTM) inspired by the latent Dirichlet allocation (LDA) model. The proposed TiCTM approach takes into account the time dependency structure between notes, measurement, and survival time of a sepsis patient. Experimental results on the public MIMIC-III database show that, overall, our method outperforms the conventional LDA and linear regression model in terms of recall, precision, accuracy, and F1-measure. It is also found that our method achieves the best performance by using 5 topics when predicting the probability for 30-day survival time.

Emotion recognition is challenging due to the emotional gap between emotions and audio–visual features. Motivated by the powerful feature learning ability of deep neural networks, this paper proposes to bridge the emotional gap by using a hybrid deep model, which first produces audio–visual segment features with Convolutional Neural Networks (CNNs) and 3D-CNN, then fuses audio–visual segment features in a Deep Belief Networks (DBNs). The proposed method is trained in two stages. First, CNN and 3D-CNN models pre-trained on corresponding large-scale image and video classification tasks are fine-tuned on emotion recognition tasks to learn audio and visual segment features, respectively. Second, the outputs of CNN and 3D-CNN models are combined into a fusion network built with a DBN model. The fusion network is trained to jointly learn a discriminative audio–visual segment feature representation. After average-pooling segment features learned by DBN to form a fixed-length global video feature, a linear Support Vector Machine is used for video emotion classification. Experimental results on three public audio–visual emotional databases, including the acted RML database, the acted eNTERFACE05 database, and the spontaneous BAUM-1s database, demonstrate the promising performance of the proposed method. To the best of our knowledge, this is an early work fusing audio and visual cues with CNN, 3D-CNN, and DBN for audio–visual emotion recognition.

Compared with ordinary single exposure images, multi-exposure fusion (MEF) images are prone to color imbalance, detail information loss and abnormal exposure in the process of combining multiple images with different exposure levels. In this paper, we proposed a human visual perception-based multi-exposure fusion image quality assessment method by considering the related perceptual features (i.e., color, dense scale invariant feature transform (DSIFT) and exposure) to measure the quality degradation accurately, which is closely related to the symmetry principle in human eyes. Firstly, the L1 norm of chrominance components between fused images and the designed pseudo images with the most severe color attenuation is calculated to measure the global color degradation, and the color saturation similarity is added to eliminate the influence of color over-saturation. Secondly, a set of distorted images under different exposure levels with strong edge information of fused image is constructed through the structural transfer, thus DSIFT similarity and DSIFT saturation are computed to measure the local detail loss and enhancement, respectively. Thirdly, Gauss exposure function is used to detect the over-exposure or under-exposure areas, and the above perceptual features are aggregated with random forest to predict the final quality of fused image. Experimental results on a public MEF subjective assessment database show the superiority of the proposed method with the state-of-the-art image quality assessment models.

Speech emotion recognition is challenging because of the affective gap between the subjective emotions and low-level features. Integrating multilevel feature learning and model training, deep convolutional neural networks (DCNN) has exhibited remarkable success in bridging the semantic gap in visual tasks like image classification, object detection. This paper explores how to utilize a DCNN to bridge the affective gap in speech signals. To this end, we first extract three channels of log Mel-spectrograms (static, delta, and delta delta) similar to the red, green, blue (RGB) image representation as the DCNN input. Then, the AlexNet DCNN model pretrained on the large ImageNet dataset is employed to learn high-level feature representations on each segment divided from an utterance. The learned segment-level features are aggregated by a discriminant temporal pyramid matching (DTPM) strategy. DTPM combines temporal pyramid matching and optimal Lp-norm pooling to form a global utterance-level feature representation, followed by the linear support vector machines for emotion classification. Experimental results on four public datasets, that is, EMO-DB, RML, eNTERFACE05, and BAUM-1s, show the promising performance of our DCNN model and the DTPM strategy. Another interesting finding is that the DCNN model pretrained for image applications performs reasonably good in affective speech feature extraction. Further fine tuning on the target emotional speech datasets substantially promotes recognition performance.

Sparse coding is currently an active subject in signal processing, computer vision, pattern recognition, etc. Fisher discrimination dictionary learning (FDDL) is a recently developed discriminative dictionary learning method and exhibits promising performance for classification. However, FDDL could not capture the locality structure of data, and it produces discriminative sparse coding coefficients, which is not effective enough for classification. To address these issues, this paper proposes an advanced version of FDDL by integrating data locality and group Lasso regularization in the procedure of FDDL's sparse coding. The proposed method is used to learn locality- and group-sensitive discriminative dictionary for facial expression recognition. Our experimental results on two public facial expression databases, i.e., the JAFFE database and the Cohn–Kanade database, demonstrate the effectiveness of the proposed method on facial expression recognition tasks, giving a significant performance improve...

Sparse coding is currently an active topic in signal processing and pattern recognition. Meta Face Learning (MFL) isatypical sparse coding method and exhibits promising performance for classification. Unfortunately, due to using the l1-norm minimization, MFLis expensive to compute and is not robust enough. To address these issues, this paper proposesafaster and more robust version of MFLwith the l2-norm regularization constraint on coding coefficients. The proposed method is used to learnaclass-specific dictionary for facial expression recognition. Extensive experiments on two popular facial expression databases, i.e., the JAFFEdatabase and the Cohn-Kanade database, demonstrate that our method shows promising computational efficiency and robustness on facial expression recognition tasks.

PM2.5-attributable deaths and years of life lost (YLL) due to specific causes during 1998-2015 in India were estimated using the integrated exposure-response (IER) model. The estimated PM2.5-mortality in India revealed an annual increasing rate of 2.7% during the study period. Spatially, deaths due to the exposure to ambient PM2.5 concentrated mostly in populated North India, and four northern states contributed 43% to the national PM2.5-attributable deaths in 2015. PM2.5-attributable deaths in India increased by 21% during 1998-2015 due to the changes of PM2.5 only, and deaths due to lung cancer (LC) revealed the largest sensitivity to increasing ambient PM2.5. The findings of this study suggest that aggressive air pollution control strategies should be implemented in North India due to their dominant contributions to the current health risks. Moreover, the rapid growth of LC related deaths with increasing ambient PM2.5 should not be neglected.

Facial expression recognition (FER) is currently a very active research topic in the fields of computer vision, pattern recognition, artificial intelligence, and has drawn extensive attentions owing to its potential applications to natural human–computer interaction (HCI), human emotion analysis, interactive video, image indexing and retrieval, etc. This paper is a survey of FER addressing the most two important aspects of designing an FER system. The first one is facial feature extraction for static images and dynamic image sequences. The second one is facial expression classification. Conclusions and future work are finally discussed in the last section of this survey.

Manifold learning and Sparse Representation Classifier are two popular techniques for face recognition. Because manifold learning can find low-dimensional representations for high-dimensional data, it is widely applied in computer vision and pattern recognition. Most of the manifold learning algorithms can be unified in the graph embedding framework, where the first step is to determine the adjacent graphs. Traditional methods use k nearest neighbor or the -ball schemes. However, they are parametric and sensitive to noises. Moreover, it is hard to determine the size of appropriate neighborhoods. To deal with these problems, in this paper, Graph Regularized Sparsity Discriminant Analysis, GRSDA, for short, is proposed. Based on graph embedding and sparsity preserving projection, the weight matrices for intrinsic and penalty graphs are obtained through sparse representation. GRSDA seeks a subspace in which samples in intra-classes are as compact as possible while samples in inter-classes are as separable as possible. Specifically, samples in the low-dimensional space can preserve the sparse locality relationship in the same class, while enhancing the separability for samples in different classes. Hence, GRSDA can achieve better performance. Extensive experiments were carried out on ORL, YALE-B and AR face databases, and the results confirmed that the proposed algorithm outperformed LPP, UDP, SPP and DSNPE.

Tissue engineering has progressed tremendously over recent decades through the generation of functional tissue analogs. Traditional approaches based on seeding cells into scaffold are limited in their capacity to produce tissues with precise biomimetic properties. Three-dimensional (3D) bioprinting is one kind of fabrication technology used to precisely dispense cell-laden biomaterials for the construction of functional tissues or organs. In recent years, much research progress has been made in 3D bioprinting technology and its application in generating tissue analogs, including skin, heart valves, blood vessels, bone, and cardiac tissue. However, it still faces many technical challenges. In this review, we introduce the current progress in 3D bioprinting technology and focus on biomaterials and their potential applications in regenerative medicine and drug discovery. Current challenges are also discussed.

In article number https://doi.org/10.1002/advs.201901822, King Wai Chiu Lai, Zhifeng Huang, Ken Kin Lam Yung, and co‐workers develop a biocompatible silica sculptured in nanozigzags by glancing angle deposition. The nanomatrix induces self‐organization of neural stem cells into miniature substantia nigra‐like brain structures, which could deliver early and progressive amelioration in Parkinsonian rats. These results provide significant insight into novel cell therapies for Parkinson's disease.

Apoptosis in intestinal epithelial cells (IECs) promotes the development of ulcerative colitis (UC), a type of inflammatory bowel disease (IBD). Efficient clearance of apoptotic cells is essential for tissue homeostasis in metazoans. Actin related protein 3 (ARP3) promotes endothelial dysfunction. The expression and function of ARP3 in UC remains unclear. In this study, the expression of apoptotic markers as p53, Bax, Cleaved-Caspease9 and Cleaved-Caspease3 were proved to be increased in the intestinal epithelial cells (IECs) of UC patients and in a mouse disuccinimidyl suberate(DSS)-induced colitis model; meanwhile, ARP3 expression was elevated. ARP3 expression levels and the severity of symptoms in patients with UC were positively correlated. By knocking down ARP3 in a TNF-α-treated NCM-460 cell colitis model, the apoptotic markers described above were all decreased. In conclusion, our data indicates that ARP3 might promote the apoptosis of IECs in UC, revealing a potential molecular target for treating UC.

UE-to-Network relay enables the user equipments (UEs) to provide relay assistance for other UEs located in close proximity. Thus, UEs, especially cell-edge UEs can obtain a better connectivity experience. In this paper, we consider the scenario where UEs can communicate with base stations (BSs) in the direct mode or in the UE-to-Network relay mode. We propose a tractable analytical framework for a UE-to-Network relay enabled uplink cellular network and investigate a flexible mode selection scheme. We leverage stochastic geometry to study the transmission power consumption, relay probability and outage probability. The analysis reveals that UE-to-Network relay is capable of reducing power consumption. Simulation results show the accuracy of our theoretical analysis and demonstrate that the proposed framework can reduce the outage probability. As a result, this paper provides guidelines for mode selection scheme in UE-to-Network relay enabled networks.

Caching popular contents at mobile devices can improve the system performance and relieve the traffic burden for the device-to-device (D2D) enabled mobile ultra dense network (UDN). In this paper, the D2D enabled mobile networks are formulated as a 3-layer architecture involving the link layer, social tie layer and preference layer. Based on the 3-layer architecture, a proactive content sharing scheme including content caching, content diffusion and resource allocation is proposed. For content caching, we propose a joint seed D2D user (DUE) and caching content selection algorithm (JSCSA) with graph theory and caching utility. With contents cached at seed DUEs, a content diffusion approach is proposed based on the social tie and D2D service priority, where seed DUEs can proactively deliver contents to other DUEs with strong social tie when seed DUEs have no requesters to serve. Furthermore, a joint subchannel allocation and power control algorithm (JSAPCA) is proposed to optimize the system throughput. Simulations are carried out to illustrate that the proposed proactive content sharing scheme has superior performances compared with the existing works.

The logistics industry is one of the major fossil energy consumers and CO2 emitters in China, which plays an important role in achieving sustainable development as well as China’s emission reduction targets. To identify the key influencing factors regarding the logistics of CO2 reductions and ensure that the development of China’s logistics industry becomes less dependent on CO2 emissions, this paper built an extended log-mean Divisia index model (LMDI) to decompose the logistics of CO2 changes between 1985 and 2015. Then, we introduced a decoupling model that combined the decomposition results to analyze the decoupling state and identify the main factors that influenced the decoupling relationship. The results show the following. (1) The urbanization effect was the decisive factor in CO2 emissions increases, followed by structural adjustment effects, while technological progress effects played a major role in inhibiting CO2 emissions. Particularly, the energy structure showed great potential for CO2 emissions reduction in China. (2) Highways appeared to have dominant promoting roles in increasing CO2 emissions regarding transportation structure effects; highways and aviation proved to have the largest impact on CO2 emission reduction. (3) There has been an increase in the number of expansive negative decoupling states between 2005 and 2015, which implies that the development of the logistics industry has become more dependent on CO2 emissions. Finally, this paper puts forward some policy implications for CO2 emission reductions in China’s logistics industry.

Deep learning is a newly-emerged machine learning theory, and has received extensive attentions in pattern recognition, signal processing, computer vision, etc. Deep belief networks (DBNs) is a representative method of deep learning and has a strong ability of unsupervised feature learning. In this paper, by combining DBNs with multi-layer perceptron (MLP), a new method of facial expression recognition based on deep learning is proposed. The proposed method integrates the DBNs's advantage of unsupervised feature learning with the MLP's classification advantage. Experimental results on two benchmarking facial expression databases, i.e., the JAFFE database and the Cohn-Kanade database, demonstrate the promising performance of the proposed method for facial expression recognition, outperforming the other state-of-the-art classification methods such as the nearest neighbour, MLP, support vector machine, the nearest subspace, as well as sparse representation-based classification.

UE-to-Network relay leverages the proximity communication between user equipments (UEs) and allows certain UEs to provide relay assistance for others, which can greatly improve system energy efficiency. In this paper, we consider the scenario where UEs suffering from bad channel condition and low battery level can communicate with the base station directly or via the help of other UEs in heterogeneous networks. The optimal power allocation and connectivity among UEs are studied, which aims at minimizing the system transmission energy while guaranteeing the minimum data rate requirement of each UE. An optimization framework is presented to formulate the system transmission energy minimization problem, which can be converted into a weighted one-to-one matching problem. And a practical joint transmission mode with relay selection and power allocation (JMRP) algorithm is developed to solve it. Simulation results show that the proposed algorithm outperforms the existing works in terms of system transmission energy and throughput.

DCAMKL1 (doublecortin and CaM kinase-like 1) has been found to be overexpressed and function as an oncogene in several types of cancer, but there are limited reports on the role of DCAMKL1 in bladder cancer. The messenger RNA and protein expression of DCAMKL1 in bladder cancer tissues and cell lines was measured by quantitative reverse transcription polymerase chain reaction, western blotting, or immunohistochemistry. The correlation between DCAMKL1 protein expression and clinicopathological characteristics was analyzed. Univariate and multivariate Cox regression models were adopted to evaluate prognostic significance of DCAMKL1 in bladder cancer patients. In our results, DCAMKL1 messenger RNA and protein were overexpressed in bladder cancer tissues compared with adjacent normal tissues. DCAMKL1 protein overexpression was positively associated with clinical stage, muscularis invasion, lymph node metastasis, and distant metastasis. The univariate and multivariate analyses suggested DCAMKL1 protein overexpression was an unfavorable prognostic factor in bladder cancer patients. In conclusion, DCAMKL1 is an independent poor prognostic factor for bladder cancer patients.

The problem of high similarity in homework has troubled teachers with time. Previous plagiarism detection systems are mainly realized by string matching which has a limitation, i.e., image homework cannot be detected. To this issue, we propose a new method of plagiarism detection in homework. First, we get fingerprint features of image homework by converting text homework into images. Then, we use image hashing algorithm and hamming distance to calculate the similarity of these features. Finally, we perform the empirical study on course of Computer Network Experiment, the test shows that our method not only reliably keeps the detection speedily, but also consistently ensures precision and false positive rate.

A new classification method called LS-KSRC is proposed.LS-KSRC integrates both sparsity and data locality in the kernel feature space.LS-KSRC's closed form solution of the l1-norm minimization problem is presented.LS-KSRC outperforms KSRC, SRC, LLC, SVM, NN, and NS. In this paper a new classification method called locality-sensitive kernel sparse representation classification (LS-KSRC) is proposed for face recognition. LS-KSRC integrates both sparsity and data locality in the kernel feature space rather than in the original feature space. LS-KSRC can learn more discriminating sparse representation coefficients for face recognition. The closed form solution of the l1-norm minimization problem for LS-KSRC is also presented. LS-KSRC is compared with kernel sparse representation classification (KSRC), sparse representation classification (SRC), locality-constrained linear coding (LLC), support vector machines (SVM), the nearest neighbor (NN), and the nearest subspace (NS). Experimental results on three benchmarking face databases, i.e., the ORL database, the Extended Yale B database, and the CMU PIE database, demonstrate the promising performance of the proposed method for face recognition, outperforming the other used methods.

Spoken emotion recognition is currently a very active research topic and has attracted extensive attention in signal processing, pattern recognition, artificial intelligence, etc. In this paper, a new emotion classification method based on kernel sparse representation, named locality-constrained kernel sparse representation-based classification (LC-KSRC), is proposed for spoken emotion recognition. LC-KSRC is able to learn more discriminating sparse representation coefficients for spoken emotion recognition, since it integrates both sparsity and data locality in the kernel feature space. The proposed method is compared with six representative emotion classification methods, including linear discriminant classifier, K-nearest-neighbor, radial basis function neural networks, support vector machines, sparse representation-based classification and kernel sparse representation-based classification. Experimental results on two publicly available emotional speech databases, i.e., the Berlin database and the Polish database, demonstrate the promising performance of the proposed method on spoken emotion recognition tasks, outperforming the other used methods.

Emotion recognition in speech signals is currently a very active research topic and has attracted much attention within the engineering application area. This paper presents a new approach of robust emotion recognition in speech signals in noisy environment. By using a weighted sparse representation model based on the maximum likelihood estimation, an enhanced sparse representation classifier is proposed for robust emotion recognition in noisy speech. The effectiveness and robustness of the proposed method is investigated on clean and noisy emotional speech. The proposed method is compared with six typical classifiers, including linear discriminant classifier, K-nearest neighbor, C4.5 decision tree, radial basis function neural networks, support vector machines as well as sparse representation classifier. Experimental results on two publicly available emotional speech databases, that is, the Berlin database and the Polish database, demonstrate the promising performance of the proposed method on the task of robust emotion recognition in noisy speech, outperforming the other used methods.

Emotion recognition is a challenging task because of the emotional gap between subjective emotion and the low-level audio-visual features. Inspired by the recent success of deep learning in bridging the semantic gap, this paper proposes to bridge the emotional gap based on a multimodal Deep Convolution Neural Network (DCNN), which fuses the audio and visual cues in a deep model. This multimodal DCNN is trained with two stages. First, two DCNN models pre-trained on large-scale image data are fine-tuned to perform audio and visual emotion recognition tasks respectively on the corresponding labeled speech and face data. Second, the outputs of these two DCNNs are integrated in a fusion network constructed by a number of fully-connected layers. The fusion network is trained to obtain a joint audio-visual feature representation for emotion recognition. Experimental results on the RML audio-visual database demonstrates the promising performance of the proposed method. To the best of our knowledge, this is an early work fusing audio and visual cues in DCNN for emotion recognition. Its success guarantees further research in this direction.

In order to study the relationship between transportation carbon emissions and economic growth, driving factors of carbon emissions, we established elastic decoupling model and LMDI decomposition analysis model of transport sector. Taking Beijing Tianjin Hebei region as an example, the empirical study was carried out.Results showed that Tianjin and Hebei were in weak decoupling state, Beijing was in expanded negative decoupling state. And the transportation output scale effect in Beijing, Tianjin and Hebei was the main factor to boost the growth of carbon emissions.Energy intensity effect has led to the reduction of carbon emissions in the transport sector in Tianjin and Hebei, but has led to the increase of transport carbon emissions in Beijing.In terms of the energy structure effect, the transportation energy structure effect of Beijing, Tianjin and Hebei have weak positive contributions or negative contribution to the overall of carbon emissions. Introduction Resolving the contradiction between economic growth and carbon emissions properly is an important task for all countries.The establishment of low carbon transportation system has an important effect on China's response to climate change and carbon emission reduction targets [1]. Therefore, it is urgent and significant to study the decoupling relationship between carbon emissions and economic growth and driving factors of carbon emissions in China's transportation industry. Decoupling theory is used to measure the relationship between material consumption and economic growth in the process of industrial development. Tapio [1] put forward the theory of elasticity decoupling. The decoupling of carbon emissions and economic growth in the transportation industry was decoupled from the economic growth. Yang [2] used the Tapio elastic decoupling model to analyze the relationship between carbon emissions and economic growth in China's transportation industry.In terms of application and study of carbon emissions structure decomposition, Ma Yueyue et al [3] used LMDI decomposition method to analyze the influence factors of carbon emissions in China logistics industry.But little research focuses on decoupling relationship between carbon emissions and economic growth, and driving factors of carbon emission of transportation industry in Beijing, Tianjin and Hebei. Beijing, Tianjin and Hebei region is one of core city circle, and also our country's political, cultural, scientific and technological center, whose transportation industry is also facing a huge energy saving and emission reduction challenges and pressure. Therefore, it is of great significance to carry out the empirical analysis on the carbon emissions decoupling effect and driving factors. Materials and Methodology In order to ensure the accuracy of carbon emissions accounting of the transportation, took 21 kinds of energy as the statistical object to calculate the carbon emissions (D) of the transport industry, see Equation (1). 44 12 i i i i D A B C N = × × × × ∑ (1) 3rd International Conference on Education, Management, Arts, Economics and Social Science (ICEMAESS 2015) © 2016. The authors Published by Atlantis Press 596 D stands for carbon emissions of energy consumption in the transport industry; Ai stands for the consumption of EnergyI; Bi stands for standard coal coefficient of EnergyI; Ci stands for carbon emission coefficient of I; Ni stands for fuel oxidation rate of Energy I. The decoupling between carbon emissions and economic growth is carbon emissions are not increased with the growth of the economy, but even appear the decline trend.Based on the Tapio elastic decoupling model, the elasticity decoupling model of the relationship between transportation industry carbon emissions and economic growth wasput forward, as shown in the formula (2). 2, 2 2 ( / ) / ( / ) CO IGDP t CO CO IGDP IGDP = D D (2) Tco2, IGDP is elasticity decoupling value of carbon emissions from transportation industry and transport industry GDP(One hundred million yuan);In order to eliminate the factor of price variation,transport industry's total production value was calculated with the constant price. This paper used extended Kaya identity and Logarithmic Mean DivisiaIndex(LMDI) to study driving factors of carbon emission. i i i i i C E E C IGDP f s n g E E IGDP = × × × = × × × ∑ ∑ (3) Ci stands for the amount of carbon emissions from the consumption of Energy I; Ei stands for the amount of consumption of Energy I; E stands for total energy consumption; IGDP stands for gross product of transport industry; fi=Ci/Ei respects for carbon intensity effect; si=Ei/E respects for the proportion of Energy I in total energy and it is used to explain the structure of energy consumption; n=E/IGDP stands for energy consumption intensity; g stands for influence of industrial scale on carbon emissions in transportation industry Using LMDI model to decompose formula (3), change of carbon emissions from the period t to t+1 period (△C)in Beijing, Tianjin and Hebei region can be expressed as the formula (4). 1 t t f s n g C C C C C C C + D = − = D + D + D + D (4)

In recent years, there is more and more debate about adherence and improve the system of tobacco monopoly or abolish tobacco monopoly system. With the increasing competition in the international tobacco market, it makes China's tobacco industry have received the unprecedented challenge. In the face of economic globalization, with the modern advanced information technology to build the tobacco industry logistics, which is the way for the survival and development in the future for China's tobacco industry. And the cost of distribution in tobacco companies occupy large proportion of the total cost of logistics, so we optimize the tobacco company logistics distribution costs based on the information environment. At last, we put forward the model of optimization of fuel management of distribution vehicle.The purpose is to reduce logistics distribution costs of tobacco companies, and improve the market competitiveness. Introduction With China entering WTO, the rising of global anti-smoking campaign and the increasing competition in the international tobacco market, it makes China's tobacco industry have received the unprecedented challenge. Under today's condition of exploring the profit is more and more difficult, the potential in the field of logistics is getting more and more attention, logistics field would be a huge "profit source". Although tobacco industry is a "monopoly" system in China, the tobacco companies are in the market economy, it is bound to be affected by the market economy. The research of logistics distribution costs which takes larger proportion in tobacco logistics’ total cost, is aimed at reducing the cost andimprovingthe competitiveness of enterprises, the research is imperative. So research of the information distribution system has practical significance for the tobacco companies. The present situation of modern tobacco logistics in China First of all, itlacksthe unified regulationsof basic equipmentin the process of goods transport. Such as the size of the container, the size of the pallet in the container, the size of the truck,the shelves size of tobacco warehouse, these all cannot be matched used. Second, the standardization of commodity information is lag. Now many of the tobacco companies are setting up their own commodity information database, but the field, type and length of the database are not consistent, this forms a lot of "information lonely islands". The exchange and sharing can't be realized in the inconsistent data, this affects the interface of tobacco logistics enterprise and the outside enterprise, also affects the operation of logistics management and e-commerce. Finally, the management of logistics standardization is backward. The traditional regional segmentation system makes the tobacco logistics divide into different departments. The lack of effective collaboration between departments leads to cost a lot of time and economic cost in the process of every mode of transportation transport of the transport links, and restricted the rapid development of logistics industry. 5th International Conference on Computer Sciences and Automation Engineering (ICCSAE 2015) © 2016. The authors Published by Atlantis Press 330 The optimization method of the logistics distribution costs Distribution costs refer to the total expenses what happened in the process of distribution.It mainly includes the distribution transportation cost, storage cost, sorting cost, equipping costs and circulation processing cost.Usually the larger cost of tobacco logistics aretwo pieces: storage and delivery.So in order to reduce the total logistics cost of the tobacco companies and enhance its market competitiveness, this paper puts forward a related optimization method to reduce the cost of logistics distribution under the environment of information technology. The optimization management for tobacco companies to the fuel consumption of distribution vehicle We can find that fuel management costs in distribution costs accounted for the proportion of larger in tobacco logistics company cost optimization analysis of the situation. It is mainly due to there is noa unified scientific consumption management method for the company.Proposed to use probability estimation algorithm to scientific management to optimize fuel consumption.Through this method to see the consumption level of the company, And by making reasonable consumption quota to manage distribution vehicle fuel consumption optimization of tobacco companies. The basic principle of probability estimation method The method of probability estimation is used in human resource management, we can introduceit into the fuel management optimization process. Probability assessment work is to further improve the experience for quality and reduce the error, and using the method of probability, to improve the accuracy of the assessment work. Probability estimation method is known as the "three valuation method", It is estimated in advance by the management of this company or experts for a working hours, the results are three kinds of numerical: Advanced working hours (To), Conservative working hours (TP), The most likely working hours (TM), then according to the following formula to calculate the advanced and reasonable working hours. T=M+λσ In an expression: T——Process planning M——Average time to process λ——The multiple of standard deviation σ ——Standard deviation of working hours of normal distribution Among them: 4 6 o p m T T T M + + = 6 p o T T σ − = 2.1.2The management optimization for vehicle fuel consumption Application the principle of probability estimation method in the fuel management optimization, the optimization steps can be as follows: (1)Statistical analysis is carried out on the thyme oil consumption situation (2)Application probability estimation method for the optimization According to the principle of probability estimation method,takethe thyme oil consumption into probability estimation,as follows: a----Advanced consumption(the consumption under the most favorable conditions.) b----Conservative consumption(the consumption under the most unfavorable conditions.) c----Normal consumption We get the above three conditions of fuel consumption value, through the analysis of the consumption data and estimates of most experienced staff. Application of probability estimation method to calculate: According to a, b, c three kinds of fuel consumption value, and obtainits average fuel consumption M average valueM=(a+4b+c)/6

Speech emotion recognition is an interesting and challenging subject due to the emotion gap between speech signals and high-level speech emotion. To bridge this gap, this paper present a method of Chinese speech emotion recognition using Deep belief networks (DBN). DBN is used to perform unsupervised feature learning on the extracted low-level acoustic features. Then, Multi-layer Perceptron (MLP) is initialized in terms of the learning results of hidden layer of DBN, and employed for Chinese speech emotion classification. Experimental results on the Chinese Natural Audio-Visual Emotion Database (CHEAVD), show that the presented method obtains a classification accuracy of 32.80 % and macro average precision of 41.54 % on the testing data from the CHEAVD dataset on speech emotion recognition tasks, significantly outperforming the baseline results provided by the organizers in the speech emotion recognition sub-challenges.

To effectively handle speech data lying on a nonlinear manifold embedded in a high-dimensional acoustic space, in this paper, an adaptive supervised manifold learning algorithm based on locally linear embedding (LLE) for nonlinear dimensionality reduction is proposed to extract the low-dimensional embedded data representations for phoneme recognition. The proposed method aims to make the interclass dissimilarity maximized, while the intraclass dissimilarity minimized in order to promote the discriminating power and generalization ability of the low-dimensional embedded data representations. The performance of the proposed method is compared with five well-known dimensionality reduction methods, i.e., principal component analysis, linear discriminant analysis, isometric mapping (Isomap), LLE as well as the original supervised LLE. Experimental results on three benchmarking speech databases, i.e., the Deterding database, the DARPA TIMIT database, and the ISOLET E-set database, demonstrate that the proposed method obtains promising performance on the phoneme recognition task, outperforming the other used methods.

Recently, compressive sensing (CS) has attracted increasing attention in the areas of signal processing, computer vision and pattern recognition. In this paper, a new method based on the CS theory is presented for robust facial expression recognition. The CS theory is used to construct a sparse representation classifier (SRC). The effectiveness and robustness of the SRC method is investigated on clean and occluded facial expression images. Three typical facial features, i.e., the raw pixels, Gabor wavelets representation and local binary patterns (LBP), are extracted to evaluate the performance of the SRC method. Compared with the nearest neighbor (NN), linear support vector machines (SVM) and the nearest subspace (NS), experimental results on the popular Cohn-Kanade facial expression database demonstrate that the SRC method obtains better performance and stronger robustness to corruption and occlusion on robust facial expression recognition tasks.

To improve effectively the performance on spoken emotion recognition, it is needed to perform nonlinear dimensionality reduction for speech data lying on a nonlinear manifold embedded in a high-dimensional acoustic space. In this paper, a new supervised manifold learning algorithm for nonlinear dimensionality reduction, called modified supervised locally linear embedding algorithm (MSLLE) is proposed for spoken emotion recognition. MSLLE aims at enlarging the interclass distance while shrinking the intraclass distance in an effort to promote the discriminating power and generalization ability of low-dimensional embedded data representations. To compare the performance of MSLLE, not only three unsupervised dimensionality reduction methods, i.e., principal component analysis (PCA), locally linear embedding (LLE) and isometric mapping (Isomap), but also five supervised dimensionality reduction methods, i.e., linear discriminant analysis (LDA), supervised locally linear embedding (SLLE), local Fisher discriminant analysis (LFDA), neighborhood component analysis (NCA) and maximally collapsing metric learning (MCML), are used to perform dimensionality reduction on spoken emotion recognition tasks. Experimental results on two emotional speech databases, i.e. the spontaneous Chinese database and the acted Berlin database, confirm the validity and promising performance of the proposed method.

Facial expression recognition is an interesting and challenging subject. Considering the nonlinear manifold structure of facial images, a new kernel-based manifold learning method, called kernel discriminant isometric mapping (KDIsomap), is proposed. KDIsomap aims to nonlinearly extract the discriminant information by maximizing the interclass scatter while minimizing the intraclass scatter in a reproducing kernel Hilbert space. KDIsomap is used to perform nonlinear dimensionality reduction on the extracted local binary patterns (LBP) facial features, and produce low-dimensional discrimimant embedded data representations with striking performance improvement on facial expression recognition tasks. The nearest neighbor classifier with the Euclidean metric is used for facial expression classification. Facial expression recognition experiments are performed on two popular facial expression databases, i.e., the JAFFE database and the Cohn-Kanade database. Experimental results indicate that KDIsomap obtains the best accuracy of 81.59% on the JAFFE database, and 94.88% on the Cohn-Kanade database. KDIsomap outperforms the other used methods such as principal component analysis (PCA), linear discriminant analysis (LDA), kernel principal component analysis (KPCA), kernel linear discriminant analysis (KLDA) as well as kernel isometric mapping (KIsomap).