Loop closure can effectively eliminate the accumulated error and plays an important role in Simultaneous Localization Mapping (SLAM). There remains challenges loop detection correction due to large viewpoints difference environment appearance changes. In this letter, we propose a novel method based on object modeling semantic graph matching. We use voxels cuboids model object-level features environment, further represent as with topological information. On basis, efficient matching edit...

Objects in environments are more stable than feature points, which can benefit for building a wide range of common view relationship and effectively reduce the cumulative error visual localization. In this article, we propose contour-SLAM, first RGB-D object-level SLAM method that uses projection constraints between dense object model its instance masks state estimation. First, considering unstructured models difficult to describe mathematically their reprojection errors quantify, use voxels...

Accurate trajectory prediction of vehicles is essential for reliable autonomous driving. To maintain consistent performance as a vehicle driving around different cities, it crucial to adapt changing traffic circumstances and achieve lifelong model. realize it, catastrophic forgetting main problem be addressed. In this paper, divergence measurement method based on conditional Kullback-Leibler proposed first evaluate spatiotemporal dependency difference among varied circumstances. Then...

Angular acceleration is an important parameter for status monitoring and fault diagnosis of rotary machinery. Therefore, we developed a novel permanent magnetic angular sensor, which without rotation angle limitations could directly measure the instantaneous rotating system. The sensor rotor only needs to be coaxially connected with system, enables convenient installation. For cup structure rotor, it has relatively small rotational inertia. Due unique mechanical output signal can directed...

To guarantee quality of service (QoS), the requirements for video transmission such as delay and cell loss rate (CLR) are very stringent. These constraints difficult to meet if high network utilization is desired. Dynamic bandwidth allocation thus needed. Video traffic prediction can play an important role in dynamic management speed networks. We propose a fast convergent algorithm predict variation I frames, based on which assigned. The new achieve convergence small error.

An angular acceleration sensor can be used for the dynamic analysis of human and joint motions. In this paper, an with novel structure based on principle electromagnetic induction is designed. The method involves construction a constant magnetic field by excitation windings sensor, cup-shaped rotor that cut field. output generate electromotive force, which directly proportional to through coupling when has rotational acceleration. mechanical working circuit are described. properties...

With rapidly increasing cyber attacks, network security has become an important issue. To protect ourselves against the Intrusion Detection System (IDS) been introduced. In such systems, different kinds of machine learning algorithms play a more and role, as support vector machine(SVM), artificial neural network(ANN), etc. False positive rate false negative rate, in addition to accuracy, are widely used for evaluation IDSs. These indices, however, often related each other, which makes it is...

Torque is an important parameter for condition monitoring and fault diagnosis rotary machines. This paper describes a new structure differential torque sensor based on the principle of electromagnetic induction. The method involves construction pulsating flux by excitation winding sensor, torsion angle produced load converted into displacements output winding. Last, generates induction potential force, which proportional to as seen through coupling. Sensor sensitivity would be reduced...

Torque changes of mechanical equipment are important information for its status, it is a necessary way to test whether various dynamic power output reached the design standard, and also an sign product qualified. application torque sensor market prospect were introduced in this paper, status domestic foreign measuring methods discussed, including variety new structure corresponding systems, at last, prospects future development measurement put forward.

Variable bit rate (VBR) video traffic exhibits high burstiness and long range dependence properties, which, in conjunction with the stringent quality of service (QoS) requirements, pose a great challenge transporting over communication network. The authors propose QoS guaranteed bandwidth allocation for given resource utilisation. Simulation analytical results demonstrate that this scheme provides delay achieves

Medical image classification plays an essential role in disease diagnosis and clinical treatment. More more research efforts have been dedicated to the design of effective methods for medical classification. As framework, capsule network (CapsNet) can realize translation equivariance. Lots current applies networks analysis. In this paper, we propose attentive octave convolutional (AOC-Caps) AOC-Caps, AOC module is used replace traditional convolution operation. The purpose process fuse high-...

SUMMARY Rendering of rigid objects with high stiffness while guaranteeing system stability remains a major and challenging issue in haptics. Being part the haptic system, behavior human operators, represented as mechanical impedance arm, has an inevitable influence on performance. This paper first verified that arm can unconsciously be modified through imposing background forces resist unstable motions arising from external disturbance forces. Then, reliable tuning (IT) method for improving...

This study focuses on the over-fitting problem in training process of deep convolutional neural network model and poor robustness when is applied an occlusion environment. We propose a unique data augmentation method, In-and-Out. First, information variance enhanced through dynamic local operation while maintaining overall geometric structure image; compared with global our method effectively alleviates overfitting significantly improves generalization ability model. Then removal operation,...

Compressive sensing (CS) is a sampling technique designed for reducing the complexity of sparse data acquisition. In this paper, modified Orthogonal Matching Pursuit (OMP) method, called Adaptive Threshold Backtracking OMP (ATBOMP) compressive and signal reconstruction presented. Compared with standard algorithm, ATBOMP method incorporates an adaptive threshold to choose candidate set ensure support set's reliability by regularized procession. Through modification, improves probability...