No abstract

Local binary pattern on three orthogonal planes (LBP-TOP) is one of the most popular method for dynamic texture analysis and has been successfully applied to facial expression analysis. Yet an effective LBP-TOP operator highly relies on preprocessing. And, like many appearance-based approaches, this approach reserves more identity-related cues rather than expression. In this work, we propose a fully automatic approach for facial expression recognition based on points registration, localized patch extraction and LBP-TOP feature representation. The efficiency of this method is evaluated on CK+ database. Results show that the proposed method has achieved a better performance compared with existing methods.

We address 3D shape classification with partial point cloud inputs captured from multiple viewpoints around the object. Different from existing methods that perform classification on the complete point cloud by first registering multi-view capturing, we propose PointView-GCN with multi-level Graph Convolutional Networks (GCNs) to hierarchically aggregate the shape features of single-view point clouds, in order to encode both the geometrical cues of an object and their multiview relations. With experiments on our novel single-view datasets, we prove that PointView-GCN produces a more descriptive global shape feature which stably improves the classification accuracy by ∼ 5% compared to the classifiers with single-view point clouds, and outperforms the state-of-the-art methods with the complete point clouds on ModelNet40.

No abstract

No abstract

Abstract. The objective of this papcr is the study of the d\'namics of damped cable systems, which are suspended in spacc, and their resonance characteristics. Of interest is tile study of the nonlinear behavior of large amplitude forced vibrations in three dimensions. As a lirst-order nonlinear problem the forced oscillations of a system having three-degrees-of-freedom with quadratic nonlinearities is developed in order to consider the resonance characteristics of the cable and the possibility of dynamic instability. The cables tire acted t.pon by their own weight in thc perpendicular direction and a steady horizontal wind. The vibrations take place about the static position of the cables as determined by the nonlinear equilibrium equations. Preliminary to the nonlinear analysis the linear mode shapes and frequencies are determined. These mode shapes tire used as coordinate functions to form weak solutions of the nonlinear autonomous partial differential equations.In order to investigate the behavior of the cablc motion in detail, the linear and the nonlinear analyses arc discussed separately. Tile lirst part of this paper deals with tile solution to the self adjoint boundary-value problenl for small-amplitude vihrations and the determination of mode shapes and natural frequencies. The second problem dealt with in this paper is tile determination of the phenomena produced by tile primary resonance of the systenl. The method of multiple time scales is used to develop solutions for the resuhing multi-dimensiomd dynamical systcnl with quadratic nonlinearity.Numerical results for the steady state response amplitude, and their variation with external cxcitation and external detuning for various values of intcrnal detuning parameters are obtained. Saturation and jump phenomena tire also obscrved The jump phenomcnon occurs when there arc multi-valued solutions and thcrc exists a variation of kinetic cncrg.v ~lll]ong solutions.

Metallography is the study of the structure of metals and alloys. Metallographic analysis can be regarded as a detection tool to assist in identifying a metal or alloy, to evaluate whether an alloy is processed correctly, to inspect multiple phases within a material, to locate and characterize imperfections such as voids or impurities, or to find the damaged areas of metallographic images. However, the defect detection of metallography is evaluated by human experts, and its automatic identification is still a challenge in almost every real solution. Deep learning has been applied to different problems in computer vision since the proposal of AlexNet in 2012. In this study, we propose a novel convolutional neural network architecture for metallographic analysis based on a modified residual neural network (ResNet). Multi-scale ResNet (M-ResNet), the modified method, improves efficiency by utilizing multi-scale operations for the accurate detection of objects of various sizes, especially small objects. The experimental results show that the proposed method yields an accuracy of 85.7% (mAP) in recognition performance, which is higher than existing methods. As a consequence, we propose a novel system for automatic defect detection as an application for metallographic analysis.

No abstract