There is a great need to assess the harmful effects or toxicities of chemicals which man exposed. In present paper, simplified molecular input line entry specification (SMILES) representation-based string kernel, together with state-of-the-art support vector machine (SVM) algorithm, were used classify toxicity from US Environmental Protection Agency Distributed Structure-Searchable Toxicity (DSSTox) database network. this method, structure can be directly encoded by series SMILES substrings...

The driving behavior of bus drivers is related to the safety all passengers and regulation urban traffic. In order analyze relevant characteristics speed acceleration, accurate trajectories patterns are essential for driver analysis development effective intelligent public transportation. Exploiting real-time vehicle tracking, this paper develops a platform with vehicle-mounted terminals using differential global navigation satellite system (DGNSS) modules analysis. DGNSS traces were used...

Abstract Speech emotion recognition is deemed to be a meaningful and intractable issue among number of do- mains comprising sentiment analysis, computer science, pedagogy, so on. In this study, we investigate speech based on sparse partial least squares regression (SPLSR) approach in depth. We make use the method implement feature selection dimensionality reduction whole acquired features. By means exploiting SPLSR method, component parts those redundant meaningless features are lessened...

Image inpainting has a good application value in image editing, however, traditional techniques cannot complete semantic repair the case of insufficient sample resources. Deep learning neural network have powerful capabilities and can extract high-level features. These features be used to semantically fill missing regions. Ideal restoration needs maintain structural consistency texture clarity. In this paper, using GAN structure, we propose method which constrains process neighborhood loss...

Although 2-D canonical correlation analysis (2DCCA) has been proposed to reduce the computational complexity while reserving local data structure of image, learned variables 2DCCA are linear combination all original variables, which makes it hard interpret solutions and might have less generality. In this paper, we propose a sparse (S2DCCA) solve drawbacks method apply image feature extraction. The basic idea S2DCCA is impose two lasso penalties on objective function obtain sets projection...

Facial image inpainting is a challenging task because the missing region needs to be filled by new pixels with semantic information (e.g., noses and mouths). The traditional methods that involve searching for similar patches are mature but it not suitable inpainting. Recently, deep generative model-based have been able implement although results blurry or distorted. In this paper, through analyzing advantages disadvantages of two methods, we propose novel efficient method combines these...

Image repair often has errors in filling. An exemplar-based technique the advantage of filling texture and structure simultaneously, but there exist some problems with this, such as not considering overall situation inpainting, unreasonable calculation priority, random selection when are multiple candidate patches. In view these situations, this paper proposes a method screen window propagating where upper layer guides lower inpainting from top by means resolution decomposition. Screen aims...

The research of multimodal emotion recognition based on facial expressions, speech, and body gestures is crucial for oncoming intelligent human–computer interfaces. However, it a very difficult task has seldom been researched in this combination the past years. Based GEMEP Polish databases, contribution focuses trimodal from gestures, including feature extraction, fusion, classification three modalities. In particular, two novel algorithms supervised least squares multiset kernel canonical...

To avoid the environmental pollution caused by livestock manure and provide rational layout of farm, we estimated phosphorus load excretion coefficient method have developed a nutrient distribution model. The was distributed to farmlands using this model spatial analysis technology. carrying capacity farms calculated based on maximum expressed in pig for Shangjie town, China. results showed that maximum, minimum, average total about 55.97, 0.74, 12.21 13,382.90 kg respectively, 2854.79 from...

In this letter, a new sparse locality preserving projection (SLPP) algorithm is developed and applied to facial expression recognition. comparison with the original (LPP) algorithm, presented SLPP able simultaneously find intrinsic manifold of feature vectors deal selection. This realized by use l1-norm regularization in LPP objective function, which directly formulated as least squares regression pattern. We two real databases (JAFFE Ekman's POFA) testify proposed method certain experiments...

The correlation between road network accessibility and land use is mainly considered a spatial-topological relationship, thus making it difficult to reveal accurately.In consideration of the speed limit structure, this paper constructed model based on weighted spatial syntax.A analysis was then conducted compared with those relative asymmetry (RA) normalized angular choice (NACH) by regarding central district Fuzhou City as an example.Results showed that roads good RA NACH were presented...

Urban rail transit systems consume significant electrical power, with traction accounting for 50% of the total energy (from China Rail Transit Association [1]). As operational mileage and vehicle counts rise, so does consumption, underscoring importance vehicular conservation. This study evaluates impact on-board storage devices on train efficiency. Using data from Changsha Metro Line 5 incorporating literature reviews simulations, energy-saving potential was assessed. Integrating these...

Based on statistics data of livestock and poultry Guangdong Province, seven characterization methods were used to characterize potential environmental pollution risk breeding (LPB) analyzed comparatively.The results showed that the different LPB generally uneven in spatial distribution.The maximum value all Shunde District, but minimum counties.In addition, number counties exceeding safety thresholds also had large differences among LPB.The number, second was by alarm pig manure equivalent...

On single sample condition, current face recognition algorithm yields serious performance drop or even fail to work, therefore a based on LGBP and locality preserving discriminant analysis (LPDA) was proposed. Image enhancement geometric transformation were firstly introduced image for reconstruction, making reconstruct original counterpart as new training set. The feature of then adopted extraction. Afterwards LPDA used dimension reduction. Classification finally accomplished by Euclidean...

We describe a new method for simultaneous image denoising and level set-based active contour segmentation using multidimensional features. consider an to be surface embedded in Riemannian manifold. By defining metric the space, which our case includes features as well model, minimization problem space can obtained through Polyakov action framework. The resulting is solved with dual algorithm efficiency. Benefits of this include fact that it independent any artificial “running” parameters,...