检索结果-内蒙古大学图书馆

3D Scene Flow Estimation on Pseudo-LiDAR: Bridging the Gap on Estimating Point Motion

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Jiang, Chaokang Wang, Guangming Miao, Yanzi Wang, Hesheng Engineering Research Center of Intelligent Control for Underground Space Ministry of Education School of Information and Control Engineering Advanced Robotics Research Center China University of Mining and Technology Xuzhou221116 China Department of Automation Key Laboratory of System Control and Information Processing Ministry of Education Key Laboratory of Marine Intelligent Equipment and System of Ministry of Education Shanghai Engineering Research Center of Intelligent Control and Management Shanghai Jiao Tong University Shanghai200240 China

3D scene flow characterizes how the points at the current time flow to the next time in the 3D Euclidean space, which possesses the capacity to infer autonomously the non-rigid motion of all objects in the scene. The previous methods for estimating scene flow from images have limitations, which split the holistic nature of 3D scene flow by estimating optical flow and disparity separately. Learning 3D scene flow from point clouds also faces the difficulties of the gap between synthesized and real data and the sparsity of LiDAR point clouds. In this paper, the generated dense depth map is utilized to obtain explicit 3D coordinates, which achieves direct learning of 3D scene flow from 2D images. The stability of the predicted scene flow is improved by introducing the dense nature of 2D pixels into the 3D space. Outliers in the generated 3D point cloud are removed by statistical methods to weaken the impact of noisy points on the 3D scene flow estimation task. Disparity consistency loss is proposed to achieve more effective unsupervised learning of 3D scene flow. The proposed method of self-supervised learning of 3D scene flow on real-world images is compared with a variety of methods for learning on the synthesized dataset and learning on LiDAR point clouds. The comparisons of multiple scene flow metrics are shown to demonstrate the effectiveness and superiority of introducing pseudo-LiDAR point cloud to scene flow estimation. © 2022, CC BY.

关键词： Three dimensional computer graphics

Recurrent Attentional Reinforcement Learning for Machinery Fault Diagnosis

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Tang, Zhenhui Wang, Jingcheng Wu, Shunyu The Department of Automation The Key Laboratory of System Control and Information Processing Ministry of Education of China The Shanghai Engineering Research Center of Intelligent Control and Management Shanghai Jiao Tong University No.800 Dongchuan Road Shanghai200240 China The SJTU Sanya Yazhou Bay Institute of Deepsea Science and Technology Sanya572024 China The Autonomous Systems and Intelligent Control International Joint Research Center Xi’an Technological University Xi’an710021 China

Recognizing fault types of machinery system is a fundamental but challenging task in industrial application. Although remarkable progress has been attained by learning fault features and predicting the corresponded fault type with deep neural networks, the existing fault diagnosis pipelines usually learn non-discriminative information. Contextual sequence dependency between the localized time fragments is often ignored. To deal with these problem, a recurrent attention framework with reinforcement learning is proposed to find a series of informative features of local time fragments related to different faults in an iterative manner. Fault labels are further predicted on these time fragments. The proposed framework also explicitly models long-term dependencies among these attentional time fragments to capture historical fault information. Experiments on two typical machinery systems, bearing system and hydraulic system, show that the proposed framework achieves superior diagnosis performance and increases no computation burden on CPU. © 2024, The Authors. All rights reserved.

关键词： Reinforcement learning

Unsupervised Learning of 3D Scene Flow with 3D Odometry Assistance

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Wang, Guangming Feng, Zhiheng Jiang, Chaokang Wang, Hesheng Department of Automation Key Laboratory of System Control and Information Processing of Ministry of Education Key Laboratory of Marine Intelligent Equipment and System of Ministry of Education Shanghai Engineering Research Center of Intelligent Control and Management Shanghai Jiao Tong University Shanghai200240 China Engineering Research Center of Intelligent Control for Underground Space Ministry of Education School of Information and Control Engineering Advanced Robotics Research Center China University of Mining and Technology Xuzhou221116 China

Scene flow represents the 3D motion of each point in the scene, which explicitly describes the distance and the direction of each point’s movement. Scene flow estimation is used in various applications such as autonomous driving fields, activity recognition, and virtual reality fields. As it is challenging to annotate scene flow with ground truth for real-world data, this leaves no real-world dataset available to provide a large amount of data with ground truth for scene flow estimation. Therefore, many works use synthesized data to pre-train their network and real-world LiDAR data to finetune. Unlike the previous unsupervised learning of scene flow in point clouds, we propose to use odometry information to assist the unsupervised learning of scene flow and use real-world LiDAR data to train our network. Supervised odometry provides more accurate shared cost volume for scene flow. In addition, the proposed network has mask-weighted warp layers to get a more accurate predicted point cloud. The warp operation means applying an estimated pose transformation or scene flow to a source point cloud to obtain a predicted point cloud and is the key to refining scene flow from coarse to fine. When performing warp operations, the points in different states use different weights for the pose transformation and scene flow transformation. We classify the states of points as static, dynamic, and occluded, where the static masks are used to divide static and dynamic points, and the occlusion masks are used to divide occluded points. The mask-weighted warp layer indicates that static masks and occlusion masks are used as weights when performing warp operations. Our designs are proved to be effective in ablation experiments. The experiment results show the promising prospect of an odometry-assisted unsupervised learning method for 3D scene flow in real-world data. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep learning

A Performance-Based Model Recovery Anti-Windup Design for Linear systems Subject to Actuator Saturation

学校读者我要写书评

暂无评论

A Performance-Based Model Recovery Anti-Windup Design for Li...

American control Conference (ACC)

作者： Wenxin Lai Yuanlong Li Zongli Lin Department of Automation Shanghai Jiao Tong University Shanghai China Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai China Shanghai Engineering Research Center of Intelligent Control and Management Shanghai China Charles L. Brown Department of Electrical and Computer Engineering University of Virginia Char-lottesville VA U.S.A.

ISBN: (数字)9798350382655

ISBN: (纸本)9798350382662

In this paper, we investigate the problem of reference tracking for a class of linear systems subject to actuator saturation and propose a performance-based model recovery anti-windup strategy. We adopt the classic model recovery anti-windup framework and regard the states or one of the outputs of the anti-windup compensator as the tracking error, by which the difference between the unconstrained system and the saturated system is quantified. Then, instead of employing the commonly used $\mathcal{L}_system$ gain, we present the prescribed performance functions (PPFs) to characterize the real-time tracking error such that the system performance can be captured more specifically. In particular, to avoid singular issues arising from the occurrence of saturation, we modify the existing PPFs with an auxiliary system when saturation occurs. Based on these modified performance functions, we follow a modified prescribed performance control approach and design the remaining output of the anti-windup compensator. Such a design procedure is constructive, making it more promising for extension to nonlinear systems. Theoretical results establish the boundedness of all signals in the closed-loop system. Simulation results verify the effectiveness of our design strategy.

关键词： Linear systems Actuators Simulation system performance Stability analysis Real-time systems Trajectory

Interactive Multi-scale Fusion of 2D and 3D Features for Multi-object Tracking

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Wang, Guangming Peng, Chensheng Zhang, Jinpeng Wang, Hesheng Department of Automation Key Laboratory of System Control and Information Processing of Ministry of Education Key Laboratory of Marine Intelligent Equipment and System of Ministry of Education Shanghai Engineering Research Center of Intelligent Control and Management Shanghai Jiao Tong University Shanghai 200240 China XLAB The Second Academy of China Aerospace Science And Industry Corporation Beijing100854 China

Multiple object tracking (MOT) is a significant task in achieving autonomous driving. Traditional works attempt to complete this task, either based on point clouds (PC) collected by LiDAR, or based on images captured from cameras. However, relying on one single sensor is not robust enough, because it might fail during the tracking process. On the other hand, feature fusion from multiple modalities contributes to the improvement of accuracy. As a result, new techniques based on different sensors integrating features from multiple modalities are being developed. Texture information from RGB cameras and 3D structure information from Lidar have respective advantages under different circumstances. However, it’s not easy to achieve effective feature fusion because of completely distinct information modalities. Previous fusion methods usually fuse the top-level features after the backbones extract the features from different modalities. In this paper, we first introduce PointNet++ to obtain multi-scale deep representations of point cloud to make it adaptive to our proposed Interactive Feature Fusion between multi-scale features of images and point clouds. Specifically, through multi-scale interactive query and fusion between pixel-level and point-level features, our method, can obtain more distinguishing features to improve the performance of multiple object tracking. Besides, we explore the effectiveness of pre-training on each single modality and fine-tuning on the fusion-based model. The experimental results demonstrate that our method can achieve good performance on the KITTI benchmark and outperform other approaches without using multi-scale feature fusion. Moreover, the ablation studies indicates the effectiveness of multi-scale feature fusion and pre-training on single modality. Copyright © 2022, The Authors. All rights reserved.

关键词： Computer vision

Incremental Online Learning of Randomized Neural Network with Forward Regularization

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Junda Hu, Minghui Li, Ning Al-Ali, Abdulaziz Suganthan, Ponnuthurai Nagaratnam The Key Laboratory of System Control and Information Processing Ministry of Education of China China The Department of Automation Shanghai Jiao Tong University Shanghai200240 China The Shanghai Engineering Research Center of Intelligent Control and Management Shanghai200240 China The School of Electrical and Electronic Nanyang Technological University Singapore The KINDI Computing Research Center College of Engineering Qatar University Qatar

Online learning of deep neural networks suffers from challenges such as hysteretic non-incremental updating, increasing memory usage, past retrospective retraining, and catastrophic forgetting. To alleviate these drawbacks and achieve progressive immediate decision-making, we propose a novel Incremental Online Learning (IOL) process of Randomized Neural Networks (Randomized NN), a framework facilitating continuous improvements to Randomized NN performance in restrictive online scenarios. Within the framework, we further introduce IOL with ridge regularization (-R) and IOL with forward regularization (-F). -R generates stepwise incremental updates without retrospective retraining and avoids catastrophic forgetting. Moreover, we substituted -R with -F as it enhanced precognition learning ability using semi-supervision and realized better online regrets to offline global experts compared to -R during IOL. The algorithms of IOL for Randomized NN with -R/-F on non-stationary batch stream were derived respectively, featuring recursive weight updates and variable learning rates. Additionally, we conducted a detailed analysis and theoretically derived relative cumulative regret bounds of the Randomized NN learners with -R/-F in IOL under adversarial assumptions using a novel methodology and presented several corollaries, from which we observed the superiority on online learning acceleration and regret bounds of employing -F in IOL. Finally, our proposed methods were rigorously examined across regression and classification tasks on diverse datasets, which distinctly validated the efficacy of IOL frameworks of Randomized NN and the advantages of forward regularization. Copyright © 2024, The Authors. All rights reserved.

关键词：

Temperature prediction of generator carbon brush based on LSTM neural network

学校读者我要写书评

暂无评论

Temperature prediction of generator carbon brush based on LS...

2021 China Automation Congress, CAC 2021

作者： Zhang, Qing Jiao, Tingyang Ma, Yinxing Zhang, Youmin Wu, Xiaohui Wu, Pengyue Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing Xi'an University of Technology Xi'an710048 China The Faculty of Computing Science The University of Alberta EdmontonABT6G 2R3 Canada Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing Xi 'an University of Technology Xi'an710048 China Department of Mechanical Industrial and Aerospace Engineering Concordia University MontrealQCH3G 1M8 Canada Huaneng Weihai Power Generation Co. Ltd Weihai264200 China Xi 'an Thermal Power Research Institute Co. Ltd Xi'an710048 China

ISBN: (纸本)9781665426473

In order to improve the prediction accuracy of the carbon brush temperature trend of the generator, analyzing the operating state of the generator more accurately, and reduce other accidents like the unplanned outage caused by carbon brush failure of the generator in the power plant. A multi-step temperature prediction model based on the LSTM (Long Short-Term Memory) neural network is proposed. The data comes from real operating carbon brush temperature of Weihai Power Plant in ten days. Then the temperature prediction model is established to achieve an accurate prediction of the carbon brush temperature in the future day. The prediction error is stable within 0.4 °C. By comparing the predicted results of the BP model and Elman model, the error between the predicted results of each model and the actual temperature data is analyzed. By comparing the indicators and analyzing the actual curves, the results show that LSTM neural network has higher accuracy in predicting the temperature of generator carbon brush. This method is of reference significance for the accurate analysis of the operation state of the generator and the load analysis of the excitation system. © 2021 IEEE

关键词： Long short-term memory

Distributed MPC Algorithm for Linear Discrete systems with Communication Noise

学校读者我要写书评

暂无评论

Distributed MPC Algorithm for Linear Discrete Systems with C...

第40届中国控制会议

作者： Duqiao Zhao Ding Liu Linxiong Liu Xia Zhang National & Local Joint Engineering Research Center of Crystal Growth Equipment and System Integration Xian University of Technology Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing

In this paper, a distributed MPC algorithm is proposed for linear discrete systems with bounded communication noise and random communication failure. Based on the characteristics of time-varying directed graphs, the convergence of the algorithm is analyzed in the case of communication noise and random communication failure. This algorithm uses the distributed projection dual subgradient algorithm to solve the dual problem and replaces the double stochastic condition of the traditional algorithm with row stochastic condition. The exponential stability of the system is guaranteed. Finally, an example is given to illustrate the performance of the algorithm.

关键词： Distributed MPC Bounded Communication Noise Random Communication Faults Distributed Projection Dual Subgradient Row Stochastic

Pseudo-LiDAR for Visual Odometry

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Deng, Huiying Wang, Guangming Feng, Zhiheng Jiang, Chaokang Wu, Xinrui Miao, Yanzi Wang, Hesheng Advanced Robotics Research Center AI Research Institute School of Information and Control Engineering China University of Mining and Technology Xuzhou221116 China Department of Automation Key Laboratory of System Control and Information Processing of Ministry of Education Key Laboratory of Marine Intelligent Equipment and System of Ministry of Education Shanghai Engineering Research Center of Intelligent Control and Management Shanghai Jiao Tong University Shanghai200240 China

As one of the important tasks in the field of robotics and machine vision, LiDAR/visual odometry provides tremendous help for various applications such as navigation, location, etc. In the existing methods, LiDAR odometry shows superior performance, but visual odometry is still widely used for its price advantage. Conventionally, the task of visual odometry mainly rely on the input of continuous images. However, it is very complicated for the odometry network to learn the epipolar geometry information provided by the images. In this paper, the concept of pseudo-LiDAR is introduced into the odometry to solve this problem. The pseudo-LiDAR point cloud back-projects the depth map generated by the image into the 3D point cloud, which changes the way of image representation. Compared with the stereo images, the pseudo-LiDAR point cloud generated by the stereo matching network can get the explicit 3D coordinates. Since the 6 Degrees of Freedom (DoF) pose transformation occurs in 3D space, the 3D structure information provided by the pseudo-LiDAR point cloud is more direct than the image. Compared with sparse LiDAR, the pseudo-LiDAR has a denser point cloud. In order to make full use of the rich point cloud information provided by the pseudo-LiDAR, a projection-aware dense odometry pipeline is adopted. Most previous LiDAR-based algorithms sampled 8192 points from the point cloud as input to the odometry network. The projection-aware dense odometry pipeline takes all the pseudo-LiDAR point clouds generated from the images except for the error points as the input to the network. While making full use of the 3D geometric information in the images, the semantic information in the images is also used in the odometry task. The fusion of 2D-3D is achieved in an image-only based odometry. Experiments on the KITTI dataset prove the effectiveness of our method. To the best of our knowledge, this is the first visual odometry method using pseudo-LiDAR. Copyright © 2022, The Authors. A

关键词： Pipelines