With the booming installed capacity of permanent magnet synchronous generator (PMSG) for wind energy generation, the grid becomes weaker and the oscillation events are frequently observed in weak-grid-tied large-scale...
详细信息
Visual reasoning between visual image and natural language description is a long-standing challenge in computer vision. While recent approaches offer a great promise by compositionality or relational computing, most o...
详细信息
ISBN:
(数字)9781728171685
ISBN:
(纸本)9781728171692
Visual reasoning between visual image and natural language description is a long-standing challenge in computer vision. While recent approaches offer a great promise by compositionality or relational computing, most of them are oppressed by the challenge of training with datasets containing only a limited number of images with ground-truth texts. Besides, it is extremely time-consuming and difficult to build a larger dataset by annotating millions of images with text descriptions that may very likely lead to a biased model. Inspired by the majority success of webly supervised learning, we utilize readily-available web images with its noisy annotations for learning a robust representation. Our key idea is to presume on web images and corresponding tags along with fully annotated datasets in learning with knowledge embedding. We present a two-stage approach for the task that can augment knowledge through an effective embedding model with weakly supervised web data. This approach learns not only knowledge-based embeddings derived from key-value memory networks to make joint and full use of textual and visual information but also exploits the knowledge to improve the performance with knowledge-based representation learning for applying other general reasoning tasks. Experimental results on two benchmarks show that the proposed approach significantly improves performance compared with the state-of-the-art methods and guarantees the robustness of our model against visual reasoning tasks and other reasoning tasks.
This paper focuses on the study of fractional order terminal sliding mode control for nonlinear aerospace systems. Firstly, a novel fractional order integral terminal sliding mode control (FO-I-TSMC) method is propose...
This paper focuses on the study of fractional order terminal sliding mode control for nonlinear aerospace systems. Firstly, a novel fractional order integral terminal sliding mode control (FO-I-TSMC) method is proposed for the control of first order nonlinear system. FO-I-TSMC has three attractive advantages: i) Non-singular control law; ii) Elimination of the reaching phase; iii) Calculable finite convergence time. Furthermore, theory analysis is presented to reveal the potential advantages of the FO-I-TSMC method over its integer order counterparts. Secondly, a novel fractional order derivative integral-TSMC (FO-DI-TSMC) method is presented to deal with second order nonlinear system. Finally, FO-DI-TSMC is extended to deal with a general class of higher order control system. Simulation results are given to verify the effectiveness of the proposed methods.
3D human face modeling is one of the most popular research directions in the field of computer stereo vision. On the one hand, the complex physiological structure of human face as well as various expressions and attit...
详细信息
3D human face modeling is one of the most popular research directions in the field of computer stereo vision. On the one hand, the complex physiological structure of human face as well as various expressions and attitude changes bring great challenges to three-dimensional modeling, which makes the modeling method of human face have high research value and reference significance. On the other hand, 3D face reconstruction has a broad application prospect, and the application demand attracts more researchers to invest in it. Based on the 3DMM and 3DDFA method, this paper utilizes regression neural network to realize end-to-end face reconstruction. After that, we build a visualization program with the training model to realize face modeling based on a single face image of any pose or expression. It’s a 3D face reconstruction system which has functions such as face alignment, face rotation and so on.
This paper aims to establish the simplest human walking model and provide a guide for controlling biped ***,a human motion capture system was utilized to sample the motion data of human walking and a representative su...
详细信息
This paper aims to establish the simplest human walking model and provide a guide for controlling biped ***,a human motion capture system was utilized to sample the motion data of human walking and a representative subject was ***,one of the simplest human walking model was set *** mechanical property analysis of human walking was done based on the established walking model via ***,the feasibility and effectiveness of proposed walking model were validated by numerical ***,the intrinsical relationships among human walking,biped walking model and control were discussed to provide a guide and insight for controlling biped robots.
The demand for stereo garages has been an urgent issue as the number of vehicles in China has been soaring these years. An intelligentcontrol system for networked stereo garages is introduced here. Based on authors’...
详细信息
With the rapid development of the stereo garage, there are plenty of security problems appeared in stereo garage. The monitoring and management system of the stereo garage has attracted more and more attention. A clou...
详细信息
Fault diagnosis is a vital technique to pinpoint the machine malfunctions in manufacturing systems. In recent years, the deep learning techniques greatly improve the fault detection accuracy, but there still remain so...
详细信息
ISBN:
(数字)9781728169040
ISBN:
(纸本)9781728169057
Fault diagnosis is a vital technique to pinpoint the machine malfunctions in manufacturing systems. In recent years, the deep learning techniques greatly improve the fault detection accuracy, but there still remain some problems. If one fault is absent in the training data or the fault signal is disturbed by severe noise interference, the fault classifier may misjudge the health state. This problem limits the reliability of the fault diagnosis in real applications. In this paper, we enhance the fault diagnosis method by using Bayesian Convolutional Neural Network (BCNN). A Shannon entropy-based method is presented to quantify the prediction uncertainty. The BCNN turns the deterministic predictions to probabilistic distributions and enhances the robustness of the fault diagnosis. The uncertainty quantification method helps to indicate the wrong predictions, detect unknown faults, and discover the strong disturbances. Then, a fine-tuning strategy is applied to enhance the model performance further. The potential usability of the proposed method in monitoring the motors of 3D printers is studied. And the experiment is conducted on a motor bearing dataset provided by Case Western Reserve University. The proposed BCNN achieves 99.82% fault classification accuracy over nine health conditions. Its robustness is verified by comparing the testing accuracy with three other methods on the noisy datasets. And the uncertainty quantification method successfully detects the outlier inputs.
As for a motion control framework of robots, virtual model control（VMC） can use virtual components to create virtual forces/torques. Actually, the virtual forces/torques are generated by joint actuators when the vir...
详细信息
As for a motion control framework of robots, virtual model control（VMC） can use virtual components to create virtual forces/torques. Actually, the virtual forces/torques are generated by joint actuators when the virtual components interact between robots and environments. In this paper, a virtual model control is proposed to do the dynamic balance control of quadruped robots in trot gait. In each leg, virtual model control includes swing phase control of robots and stance phase counterparts. In whole body, based on the forces/torques distribution method between two stance legs, virtual model control is mainly about the attitude control containing roll, pitch and yaw. Then, an intuitive approach of velocity control is employed for the locomotion of quadruped robots. Based on the velocity planning and control, a trajectory tracking control approach is investigated by considering four factors: terrain complexity index, curvature radius of given trajectory, distance to terminal, and maximum velocity of quadruped robots. Finally, the effectiveness of proposed controllers is validated by co-simulations.
This study addresses the uniformly globally asymptotically stability (UGAS) problem of switched nonlinear delay systems (SNDSs) with sampled-data inputs (SDIs). By using multiple Lyapunov functionals (MLFs) method, mo...
详细信息
This study addresses the uniformly globally asymptotically stability (UGAS) problem of switched nonlinear delay systems (SNDSs) with sampled-data inputs (SDIs). By using multiple Lyapunov functionals (MLFs) method, mode-dependent average dwell times, and the total activating time length of MLFs, some stability criteria are explicitly obtained for SNDSs with SDIs. Meanwhile, the UGAS property for SNDSs with some or all unstable modes is investigated. For unstable modes and stable modes, we adopt different switching signals. Besides, we establish some sufficient stability conditions in the form of an upper bound on the sum of dwell times and sampling intervals. Simulation examples are adopted to illustrate and verify the effectiveness of our proposed methods.
暂无评论