Fall events have unique dynamic features,which are not fully utilized by existing fall detection *** on video understanding,we propose Fall-LSTM to learn such features pertinently without additional ***-LSTM is compos...
Fall events have unique dynamic features,which are not fully utilized by existing fall detection *** on video understanding,we propose Fall-LSTM to learn such features pertinently without additional ***-LSTM is composed of CNN-LSTM framework and two excitation modules i.e.,Spatial Attention Module(SAM) and Temporal Location Module(TLM).SAM provides spatial constraints on motion for feature layers through foreground extraction and spatial *** emphasizes frames with high probability of fall events to LSTM by inferring the rate and trend of motion in *** results show that our proposed modules significantly improve the performance of LSTM model,outperforming the state-of-theart methods on two public Fall Detection Datasets i.e.,Le2 i and UR.
We construct the first markerless deformable interaction dataset recording interactive motions of the hands and deformable objects, called HMDO (Hand Manipulation with Deformable Objects). Our motivation is the curren...
In this paper, the stability analysis of Load frequency control (LFC) systems with time-varying delay is conducted. Firstly, an augmented Lyapunov-Krasovskii (L-K) functional is designed to incorporate the relevant in...
详细信息
The partial shading condition(PSC) makes it challenging for the PV system to find the maximum power *** this paper,an improved gray wolf algorithm(GWO) is proposed by introducing elimination mechanism,greedy mechanism...
详细信息
The partial shading condition(PSC) makes it challenging for the PV system to find the maximum power *** this paper,an improved gray wolf algorithm(GWO) is proposed by introducing elimination mechanism,greedy mechanism and adjusting the convergence factor which overcomes the contradiction between global exploration ability and convergence *** fast varying irradiance and PSC,IGWO has been compared with GWO,particle swarm optimization(PSO),and adaptive particle swarm optimization(APSO),the results shows the superiority of IGWO in the MPPTs of the PV system.
In order to optimize the control performance of the permanent magnet synchronous motor(PMSM) in the presence of strong disturbances and model uncertainty,a composite single-loop model predictive control(MPC) based...
详细信息
In order to optimize the control performance of the permanent magnet synchronous motor(PMSM) in the presence of strong disturbances and model uncertainty,a composite single-loop model predictive control(MPC) based on disturbance observer(DOB) is studied in this ***,a DOB is employed to estimate the lumped disturbance including the uncertain parameters and the external disturbance in real ***,benefitting from model-based design,estimations are introduced into the prediction model of the PMSM,then a composite single-loop MPC is designed based on this ***,the optimal control input is obtained,by introducing Lagrange vector and solving the quadratic programming problem with *** results show that the proposed composite control method has a faster transient response and a better disturbance rejection *** addition,compared with the generally cascade structure,only the predictive horizon and the observer poles are needed to be adjusted,so it provides a more practical optimization algorithm for the engineering application.
This paper proposes a two-column convolutional neural network algorithm for face point detection, which proves better performance than single-column. Because of the deep level structure, global texture features are ex...
详细信息
This paper proposes a two-column convolutional neural network algorithm for face point detection, which proves better performance than single-column. Because of the deep level structure, global texture features are extracted at higher layers of the network, and the results of keypoint location remains high accuracy. In the first column, we take the channel component of original image as input to train the neural network respectively and calculate the average location results. In the second one, we use Sobel operator to extract first-order derivative feature of original image and regard it as the input to train another convolutional network. The location results of two columns above are fused with appropriate proportion to ensure better results. Experimental result on LFW dataset shows the performance of the proposed parallel structure better than single convolutional network, and it stays robust due to occlusions, large pose variations and extreme lighting.
Aiming at the problem that human breathing and heartbeat signals are difficult to detect in a low signal-to-clutter environment,this paper proposes a signal processing method that combines differential evolution algor...
详细信息
Aiming at the problem that human breathing and heartbeat signals are difficult to detect in a low signal-to-clutter environment,this paper proposes a signal processing method that combines differential evolution algorithm and variational modal decomposition *** order to achieve non-contact detection of vital signs,this paper uses ultra-wideband radar to collect the signal reflected by the human body,and suppresses the interference of the background signal by removing the background of the signal,and then extracts the signal reflected by the human chest in the fast time *** that,according to the frequency range of breathing and heartbeat signals,low-pass filter and band-pass filter are used to filter useless *** at the problem that the variational modal decomposition algorithm needs to manually adjust parameters in advance,this paper takes the maximum effective frequency band energy ratio as the objective function,and uses differential evolution algorithm to automatically search for the best parameter combination of variational modal decomposition,which reduces the workload of parameter ***,variational modal decomposition by the best parameter combination is used again to decompose the signal,and the breathing and heartbeat signals are reconstructed from the decomposed modal *** show that the method proposed in this paper can measure the human body's respiratory rate and heartbeat rate with high accuracy.
Natural Locomotion interface(NLI) is critical to expanding the users' exploration of scenes in virtual reality and improving user *** on the 2D motion platform,users can achieve a natural locomotion experience in ...
详细信息
Natural Locomotion interface(NLI) is critical to expanding the users' exploration of scenes in virtual reality and improving user *** on the 2D motion platform,users can achieve a natural locomotion experience in a limited physical *** a small-scale 2D motion platform,an ideal situation is that the velocity of the platform can always be synchronized with the user's actual intended velocity,so that the user's center of mass is kept at the center of the *** puts a brand new requirement on the performance of the platform *** paper designs an acceleration-level state feedback controller for the small-scale 2D motion *** the user's intended acceleration as an external disturbance,it is estimated by introducing a disturbance state observer;for the user' s velocity on the platform,a linear state observer is used to estimate it;then takes the estimated values as feedforward terms to compensate the *** with the HCMK1 2D motion platform,we implemented the controller and verified it to be *** recording the data of the user walking along the circle and the square trajectory,it was verified that the designed controller has good control performance for the user's motion and the state observers a great performance to quickly track the user's intended acceleration and the user's velocity on the platform.
Achieving optimal speed regulation for permanent magnet synchronous motors (PMSMs) remains a challenging task, particularly in selecting the most suitable controller to meet desired objectives. This paper considers th...
详细信息
Dynamic multi-person mesh recovery has been a hot topic in 3D vision recently. However, few works focus on the multi-person motion capture from uncalibrated cameras, which mainly faces two challenges: the one is that ...
详细信息
暂无评论