检索结果-内蒙古大学图书馆

Visual Explanation of Object Detectors via Saliency Maps 18

学校读者我要写书评

暂无评论

Visual Explanation of Object Detectors via Saliency Maps

18th IEEE Conference on Industrial Electronics and Applications, ICIEA 2023

作者： Xiao, Jin Liu, Wenrui Hu, Xiaoguang Jiang, Hao Wang, Weipeng Yahang, Li Beihang University State Key Laboratory of Virtual Reality Technology and Systems Beijing China Beihang University Beijing China The Third Academy of Casic Beijing China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Beijing China

ISBN: (纸本)9798350312201

In recent years, rapid progress of deep learning has resulted in object detection algorithms based on deep neural networks nearly completely replacing traditional methods. This shift has led to significant improvements in both accuracy and speed of object detection. However, deep neural networks that support decision-making inobject detection are complex black boxes with hidden internal logic and operations. This lack of transparency results in deep learning models being unexplainable, which limits their practical deployment. In this paper, we analyze the current state and future trends of interpretable technology in object detection and propose an effective object detection interpretable algorithm. This algorithm aims to explain the underlying logic of AI decision-making and provide support for the high performance and credibility of object detection algorithms based on deep neural networks through quantitative evaluation indicators. © 2023 IEEE.

关键词： Object detection

MC-NeRF: Multi-Camera Neural Radiance Fields for Multi-Camera Image Acquisition systems

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Gao, Yu Su, Lutong Liang, Hao Yue, Yufeng Yang, Yi Fu, Mengyin School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex System Beijing Institute of Technology Beijing China

Neural Radiance Fields (NeRF) use multi-view images for 3D scene representation, demonstrating remarkable performance. As one of the primary sources of multi-view images, multi-camera systems encounter challenges such as varying intrinsic parameters and frequent pose changes. Most previous NeRF-based methods assume a unique camera and rarely consider multi-camera scenarios. Besides, some NeRF methods that can optimize intrinsic and extrinsic parameters still remain susceptible to suboptimal solutions when these parameters are poor initialized. In this paper, we propose MC-NeRF, a method that enables joint optimization of both intrinsic and extrinsic parameters alongside NeRF. The method also supports each image corresponding to independent camera parameters. First, we tackle coupling issue and the degenerate case that arise from the joint optimization between intrinsic and extrinsic parameters. Second, based on the proposed solutions, we introduce an efficient calibration image acquisition scheme for multi-camera systems, including the design of calibration object. Finally, we present an end-to-end network with training sequence that enables the estimation of intrinsic and extrinsic parameters, along with the rendering network. Furthermore, recognizing that most existing datasets are designed for a unique camera, we construct a real multi-camera image acquisition system and create a corresponding new dataset, which includes both simulated data and real-world captured images. Experiments confirm the effectiveness of our method when each image corresponds to different camera parameters. Specifically, we use multi-cameras, each with different intrinsic and extrinsic parameters in real-world system, to achieve 3D scene representation without providing initial poses. The Code and supplementary materials are available at https://***/MC-NeRF Copyright © 2023, The Authors. All rights reserved.

关键词： Parameter estimation

DR-WLC: Dimensionality Reduction cognition for object detection and pose estimation by Watching, Learning and Checking

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Gao, Yu Xu, Xi Jiang, Tianji Chen, Siyuan Yang, Yi Yue, Yufeng Fu, Mengyin School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex System Beijing Institute of Technology Beijing China

Object detection and pose estimation are difficult tasks in robotics and autonomous driving. Existing object detection and pose estimation methods mostly adopt the same-dimensional data for training. For example, 2D object detection usually requires a large amount of 2D annotation data with high cost. Using high-dimensional information to supervise lower-dimensional tasks is a feasible way to reduce datasets size. In this work, the DR-WLC, a dimensionality reduction cognitive model, which can perform both object detection and pose estimation tasks at the same time is proposed. The model only requires 3D model of objects and unlabeled environment images (with or without objects) to finish the training. In addition, a bounding boxes generation strategy is also proposed to build the relationship between 3D model and 2D object detection task. Experiments show that our method can qualify the work without any manual annotations and it is easy to deploy for practical applications. Source code is at https: //***/IN2-ViAUn/DR-WLC. Copyright © 2023, The Authors. All rights reserved.

关键词： Object detection

Sector Patch Embedding: An Embedding Module Conforming to The Distortion Pattern of Fisheye Image

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yang, Dianyi Tang, Jiadong Gao, Yu Yang, Yi Fu, Mengyin School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex System Beijing Institute of Technology Beijing China

Fisheye cameras suffer from image distortion while having a large field of view(LFOV). And this fact leads to poor performance on some fisheye vision tasks. One of the solutions is to optimize the current vision algorithm for fisheye images. However, most of the CNN-based methods and the Transformer-based methods lack the capability of leveraging distortion information efficiently. In this work, we propose a novel patch embedding method called Sector Patch Embedding(SPE), conforming to the distortion pattern of the fisheye image. Furthermore, we put forward a synthetic fisheye dataset based on the ImageNet-1K and explore the performance of several Transformer models on the dataset. The classification top-1 accuracy of ViT and PVT is improved by 0.75% and 2.8% with SPE respectively. The experiments show that the proposed sector patch embedding method can better perceive distortion and extract features on the fisheye images. Our method can be easily adopted to other Transformer-based models. Source code is at https://***/IN2-ViAUn/Sector-Patch-Embedding. Copyright © 2023, The Authors. All rights reserved.

关键词： Embeddings

CRRS: Concentric Rectangles Regression Strategy for Multi-point Representation on Fisheye Images

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Xihan Xu, Xi Gao, Yu Yang, Yi Yue, Yufeng Fu, Mengyin School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex System Beijing Institute of Technology Beijing China

Modern object detectors take advantage of rectangular bounding boxes as a conventional way to represent objects. When it comes to fisheye images, rectangular boxes involve more background noise rather than semantic information. Although multi-point representation has been proposed, both the regression accuracy and convergence still perform inferior to the widely used rectangular boxes. In order to further exploit the advantages of multi-point representation for distorted images, Concentric Rectangles Regression Strategy(CRRS) is proposed in this work. We adopt smoother mean loss to allocate weights and discuss the effect of hyper-parameter to prediction results. Moreover, an accurate pixel-level method is designed to obtain irregular IoU for estimating detector performance. Compared with the previous work for muti-point representation, the experiments show that CRRS can improve the training performance both in accurate and stability. We also prove that multi-task weighting strategy facilitates regression process in this design. Source code is at https://***/IN2-ViAUn/Concentric-Rectangles-Loss. Copyright © 2023, The Authors. All rights reserved.

关键词： Semantics

Real-time terrain assessment and Bayesian-based path planning for off-road navigation

学校读者我要写书评

暂无评论

Real-time terrain assessment and Bayesian-based path plannin...

IEEE/RSJ International Conference on intelligent Robots and systems (IROS)

作者： Tianwei Niu Shuwei Yu Liang Wang Haoyu Yuan Shoukun Wang Junzheng Wang Localization and Motion Planning School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex System School of Automation Beijing Institute of Technology Beijing China

ISBN: (数字)9798350377705

ISBN: (纸本)9798350377712

In the context of unstructured and unknown environment, the autonomous navigation still faces many challenges, such as assessing rough terrain and deciding how to safely navigate complex terrain. In this work, we propose a robust and practical off-road navigation framework that has been successfully deployed on a vibroseis truck for land exploration. First, in degraded wild scenes, a tightly coupled lidar-GNSS-inertial fusion odometry and mapping framework is adopted to construct a local point cloud map around the vehicle in real-time and provide precise localization. Then, based on amplitude-frequency characteristic analysis and point cloud PCA, a multi-layer terrain assessment map containing terrain roughness, obstacles and slope information is obtained. Finally, combining Gaussian distribution based adaptive sampler and Bayesian sequentially updated proposal distribution, a local graph is efficiently built to obtain multiple path solutions under constrained conditions. Both simulations and field experiments show that the proposed navigation framework can decide how to travel on a flat road even in harsh terrain conditions, naturally suppressing frequent attitude angle changes and preventing vehicle accidents.

关键词： Point cloud compression Location awareness Adaptation models Navigation Roads Gaussian distribution Real-time systems Bayes methods Proposals Principal component analysis

Design of a High Speed 360-degree Panoramic Video Acquisition system Based on FPGA and USB 3.0

学校读者我要写书评

暂无评论

IEEE Sensors Journal 2023年 1-1页

作者： Lyu, Congyi Peng, Jianqing Zhou, Weiguo Yang, Shanshan Liu, Yunhui Intelligent Robotics Institute Key Laboratory of Biomimetic Robots and Systems Ministry of Education State Key Laboratory of Intelligent Control and Decision of Complex Systems School of Mechatronical Engineering Beijing Institute of Technology Beijing China

This paper presents a FPGA-based hardware architecture for capturing real-time panoramic images and the parallel implementation of an image stitching algorithm based on the scale-invariant feature transform (SIFT) features. First, the system transfers the images captured by five CMOS sensors to the designed FPGA board through the FMC adapter plate. Then, the system employs the FPGA-based parallel computing to extract the SIFT features and detect the best matching region and points for image registration. Finally, in order to obtain the panoramic image with smooth changes at the junctions, images of the adjacent images are fused. Using the designed FPGA platform makes it possible to extract the SIFT and stitch the images in real-time, which is not realized in most panoramic imaging systems. The panoramic system, developed using the Xilinx Zynq®-7000 all programmable SOC, can capture video of 5*640*480 images at the speed of 60 fps. To enable high-speed image transfer from the imaging system to the host computer, the USB3.0 interface with the transmission speed of 359MB/s has been developed. The real-time efficiency of the parallel system will be demonstrated by the simulations and experiments. IEEE

关键词： Real time systems

PI Tuning and Optimization for Speed control of PMSM Based on Multivariable Sliding-Mode Extremum Seeking

学校读者我要写书评

暂无评论

PI Tuning and Optimization for Speed Control of PMSM Based o...

第34届中国控制与决策会议

作者： Fan Li Wei Shen Chuanyu Ge Key Laboratory of Complex System Intelligent Control and Decision School of AutomationBeijing Institute of Technology

ISBN: (纸本)9781665478977

The PI controller is the most basic and widely used for the permanent magnet synchronous motor speed *** cumbersome parameter adjustment process makes it difficult for the PI controller to balance convenience and *** this end,this paper presents a new PI tuning and optimization scheme based on multivariable sliding-mode extremum ***,by analyzing the frequency domain model of permanent magnet synchronous motor,the analytical calculation formulas of controller parameters and motor parameters are derived as the initial value of PI parameter ***,a multivariable sliding-mode extremum seeking optimizer is designed to optimize the control parameters by minimizing the cost function based on the feedback error *** feasibility of the above scheme is proved by theoretical derivation,and the simulation and experimental results also verify the feasibility and effectiveness of the proposed method.

关键词： permanent-magnet synchronous motor(PMSM) multivariable sliding mode extremum seeking(MSME) self-tuning parameter optimization

Adaptive Temporal Aggregation Network for Online Video Visual Relation Detection

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： He, Yuxuan Gan, Ming-Gang Ma, Qianzhao Wang, Chenyu State Key Laboratory of Intelligent Control and Decision of Complex Systems School of Automation Beijing Institute of Technology Beijing100081 China

Nowadays, more and more researchers pay attention to scene perception of artificial robot. Video visual relation detection is an essential task for scene perception but existing methods are all offline methods which are unsuitable for real-time application. Thus, we first introduce online video visual relation detection task which aims at localizing and classifying relations in the current frame. Then, we propose Adaptive Temporal Aggregation Network for online video visual relation detection. In this network, we design content-based adaptive sampling method to sample historical frames which contain dynamic scene information. In order to extract history knowledge from these discrete features, we propose adaptive spatio-temporal activation module which consists of adaptive spatial activation unit and adaptive temporal aggregation unit to extract history knowledge from history memory. After obtaining history knowledge and current state, the proposed attention-based knowledge-state fusion module is used to fuse these dynamic and static visual information. We also construct spatial location features and semantic embedding features to improve final relation representations. Experiments show the effectiveness and efficiency of our method. Moreover, we implement our network on fetch robot and visualize the online video visual relation detecting process. © 2023, The Authors. All rights reserved.

关键词： Chemical activation