检索结果-内蒙古大学图书馆

2025 IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2025

作者： Witte, Christian Behley, Jens Stachniss, Cyrill Raaijmakers, Marvin Cariad Se Germany University of Bonn Center for Robotics Germany Lamarr Institute for Machine Learning and Artificial Intelligence Germany

ISBN: (纸本)9798331510831

Spatial understanding of the semantics of the surroundings is a key capability needed by autonomous cars to enable safe driving decisions. Recently, purely vision-based solutions have gained increasing research interest. In particular, approaches extracting a bird's eye view (BEV) from multiple cameras have demonstrated great performance for spatial understanding. This paper addresses the dependency on learned positional encodings to correlate image and BEV feature map elements for transformer-based methods. We propose leveraging epipolar geometric constraints to model the relationship between cameras and the BEV by Epipolar Attention Fields. They are incorporated into the attention mechanism as a novel attribution term, serving as an alternative to learned positional encodings. Experiments show that our method EAFormer outperforms previous BEV approaches by 2% mIoU for map semantic segmentation and exhibits superior generalization capabilities compared to implicitly learning the camera configuration. © 2025 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Adaptive Neural Network Dynamic Surface Trajectory Tracking Control for Underactuated Autonomous Underwater Vehicles 6

Adaptive Neural Network Dynamic Surface Trajectory Tracking ...

引用

6th International Conference on robotics and Computer Vision, ICRCV 2024

作者： Jiang, Guohong Joo Er, Meng Gong, Huibin Wang, Siqi Institute of Artificial Intelligence and Marine Robotics College of Marine Electrical Engineering Dalian Maritime University Dalian China

ISBN: (纸本)9798331527426

Aiming at the trajectory tracking control problem of underactuated autonomous underwater vehicles (AUVs) with unknown dynamic model parameters and time-varying ocean disturbances, a dynamic surface adaptive control strategy based on radial basis function neural network is proposed to complete the desired three-dimensional trajectory tracking. Firstly, an output redefinition method is used to solve the underactuated problem. Secondly, a dynamic surface control method is used to overcome the 'explosion of complexity' problem in traditional backstepping control, and complexity of the control law design is simplified. At the same time, a nonlinear damping term is added in control law to overcome unknown external disturbances. Finally, unknown dynamic model parameters are approximated by a neural network, and the neural network weights are estimated online by an adaptive law. The Lyapunov theory is used to prove that all signals of closed-loop system are semi-globally uniformly bounded, and effectiveness of the control law is verified by comparative simulations. © 2024 IEEE.

关键词： Radial basis function networks

来源：评论

学校读者我要写书评

暂无评论

Mars Planner: Improved Batch Spatio-Temporal Path Planning for Multi-Ackerman Robotic Systems 27

Mars Planner: Improved Batch Spatio-Temporal Path Planning f...

引用

27th IEEE International Conference on Intelligent Transportation Systems, ITSC 2024

作者： Guo, Song Wang, Shen'ao He, Junjie Chen, Liming Wang, Hang Sun, Hongbin School of Microelectronics Xi'an Jiaotong University Shaanxi Xi'an710049 China Institute of Artificial Intelligence and Robotics College of Artificial Intelligence Xi'an Jiaotong University Shaanxi Xi'an710049 China

ISBN: (纸本)9798331505929

This paper introduces an innovative multi-agent path finding (MAPF) system specifically designed for navigating multi-Ackerman robotic systems in intricate environments. The Mars Planner, the proposed solution, enhances path planning by tackling collision-free path challenges encountered by groups of intelligent agents. Our contributions include the development of two key algorithms: the Fast Batch Path Finding (FBPF) and the Batch Spatio-Temporal Path Refinement (BSTPR). FBPF utilizes a hybrid A∗ approach to generate preliminary coarse paths within free configuration spaces, while BSTPR refines these paths using topological homotopy strategies to optimize time allocation and effectively resolve internal conflicts. Through simulations and physical experiments, we demonstrate significant enhancements in computational efficiency and path quality compared to existing methods. In conclusion, the Mars Planner stands as an efficient solution capable of managing large-scale complexity in real-world applications. It offers a robust and scalable framework suitable for diverse environments and scenarios. © 2024 IEEE.

关键词： Intelligent agents

来源：评论

学校读者我要写书评

暂无评论

Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Model 41

Vision Mamba: Efficient Visual Representation Learning with ...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Zhu, Lianghui Liao, Bencheng Zhang, Qian Wang, Xinlong Liu, Wenyu Wang, Xinggang School of EIC Huazhong University of Science & Technology China Institute of Artificial Intelligence Huazhong University of Science & Technology China Horizon Robotics China Beijing Academy of Artificial Intelligence China

Recently the state space models (SSMs) with efficient hardware-aware designs, i.e., the Mamba deep learning model, have shown great potential for long sequence *** building efficient and generic vision backbones purely upon SSMs is an appealing ***, representing visual data is challenging for SSMs due to the position-sensitivity of visual data and the requirement of global context for visual *** this paper, we show that the reliance on self-attention for visual representation learning is not necessary and propose a new generic vision backbone with bidirectional Mamba blocks (Vim), which marks the image sequences with position embeddings and compresses the visual representation with bidirectional state space *** ImageNet classification, COCO object detection, and ADE20k semantic segmentation tasks, Vim achieves higher performance compared to well-established vision transformers like DeiT, while also demonstrating significantly improved computation & memory *** example, Vim is 2.8× faster than DeiT and saves 86.8% GPU memory when performing batch inference to extract features on images with a resolution of 1248×*** results demonstrate that Vim is capable of overcoming the computation & memory constraints on performing Transformer-style understanding for high-resolution images and it has great potential to be the next-generation backbone for vision foundation *** and models are released at https://***/hustvl/Vim. Copyright 2024 by the author(s)

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Layer-Animate for Transparent Video Generation

Layer-Animate for Transparent Video Generation

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Bai, Jingqi Zhou, Jingkai Wang, Benzhi Chen, Weihua Yang, Yang Lei, Zhen Wang, Fan State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences China School of Artificial Intelligence University of Chinese Academy of Sciences China Centre for Artificial Intelligence and Robotics Hong Kong Institute of Science & Innovation Chinese Academy of Sciences China Alibaba Group China

ISBN: (纸本)9798350368741

Transparent videos with alpha channels play a crucial role in film production, advertising, and augmented reality fields. However, there is currently no available method for producing transparent videos. Traditional methods are time-consuming and labor-intensive, and employing alternative approaches for this task will result in inaccurate transparent regions, constrained motion, and artifacts. To address these challenges, we propose Layer-Animate, the first method capable of generating transparent videos. Our method comprises two stages: in the first stage, transparent images are generated as the base images to provide content and transparency information for the next stage. In the second stage, Inter-Frame Attention is applied to decouple content from motion, enabling the motion module to focus better on action. Layer-Animate is the first method used to generate transparent videos with accurate transparent regions, sufficient motion, and no artifacts, as demonstrated by notable improvements in qualitative and quantitative metrics. © 2025 IEEE.

关键词： deep generative models diffusion models transparent videos video generation

来源：评论

学校读者我要写书评

暂无评论

DOES DEEP LEARNING LEARN TO ABSTRACT? A SYSTEMATIC PROBING FRAMEWORK 11

DOES DEEP LEARNING LEARN TO ABSTRACT? A SYSTEMATIC PROBING F...

引用

11th International Conference on Learning Representations, ICLR 2023

作者： An, Shengnan Lin, Zeqi Chen, Bei Fu, Qiang Zheng, Nanning Jian-Guang, L.O.U. Institute of Artificial Intelligence and Robotics Xi'an Jiaotong University China Microsoft Corporation China

Abstraction is a desirable capability for deep learning models, which means to induce abstract concepts from concrete instances and flexibly apply them beyond the learning context. At the same time, there is a lack of clear understanding about both the presence and further characteristics of this capability in deep learning models. In this paper, we introduce a systematic probing framework to explore the abstraction capability of deep learning models from a transferability perspective. A set of controlled experiments are conducted based on this framework, providing strong evidence that two probed pre-trained language models (PLMs), T5 and GPT2, have the abstraction capability. We also conduct in-depth analysis, thus shedding further light: (1) the whole training phase exhibits a "memorize-then-abstract" two-stage process;(2) the learned abstract concepts are gathered in a few middle-layer attention heads, rather than evenly distributed throughout the model;(3) the probed abstraction capabilities exhibit robustness against concept mutations, and are more robust to low-level/source-side mutations than high-level/target-side ones;(4) generic pre-training is critical to the emergence of abstraction capability, and PLMs exhibit better abstraction with larger model sizes and data scales. © 2023 11th International Conference on Learning Representations, ICLR 2023. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Parallel Bounded Search for the Maximum Clique Problem

引用

Journal of Computer Science & Technology 2023年第5期38卷 1187-1202页

作者：江华白珂刘海姣李初民 Felip Manya 付樟华 Engineering Research Center of Cyberspace Yunnan UniversityKunming 650500China School of Software Yunnan UniversityKunming 650500China Laboratory of Modeling Information and SystemsUniversity of Picardie Jules VerneAmiens 80039France Artificial Intelligence Research Institute Spanish National Research CouncilCatalonia 08193Spain Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen 518000China Institute of Robotics and Intelligent Manufacturing Chinese University of Hong KongShenzhen 518000China

Given an undirected graph,the Maximum Clique Problem(MCP)is to find a largest complete subgraph of the *** is NP-hard and has found many practical *** this paper,we propose a parallel Branch-and-Bound(BnB)algorithm to tackle this NP-hard problem,which carries out multiple bounded searches in *** search has its upper bound and shares a lower bound with the rest of the *** potential benefit of the proposed approach is that an active search terminates as soon as the best lower bound found so far reaches or exceeds its upper *** describe the implementation of our highly scalable and efficient parallel MCP algorithm,called PBS,which is based on a state-of-the-art sequential MCP *** proposed algorithm PBS is evaluated on hard DIMACS and BHOSLIB *** results show that PBS achieves a near-linear speedup on most DIMACS instances and a superlinear speedup on most BHOSLIB ***,we give a detailed analysis that explains the good speedups achieved for the tested instances.

关键词： Branch-and-Bound(BnB) maximum clique problem(MCP) parallel search

来源：评论

学校读者我要写书评

暂无评论

Human-Like Decision-Making of Autonomous Vehicles in Dynamic Traffic Scenarios

引用

IEEE/CAA Journal of Automatica Sinica 2023年第10期10卷 1905-1917页

作者： Tangyike Zhang Junxiang Zhan Jiamin Shi Jingmin Xin Nanning Zheng National Key Laboratory of Human-Machine Hybrid Augmented Intelligence National Engineering Research Center for Visual Information and ApplicationsXi’an 710049 Institute of Artificial Intelligence and Robotics Xi’an Jiaotong UniversityXi’an 710049China Momenta Shanghai 201804China

With the maturation of autonomous driving technology, the use of autonomous vehicles in a socially acceptable manner has become a growing demand of the public. Human-like autonomous driving is expected due to the impact of the differences between autonomous vehicles and human drivers on *** human-like decision-making has become a research hotspot, a unified theory has not yet been formed, and there are significant differences in the implementation and performance of existing methods. This paper provides a comprehensive overview of human-like decision-making for autonomous vehicles. The following issues are discussed: 1) The intelligence level of most autonomous driving decision-making algorithms;2) The driving datasets and simulation platforms for testing and verifying human-like decision-making;3) The evaluation metrics of human-likeness;personalized driving;the application of decisionmaking in real traffic scenarios;and 4) The potential research direction of human-like driving. These research results are significant for creating interpretable human-like driving models and applying them in dynamic traffic scenarios. In the future, the combination of intuitive logical reasoning and hierarchical structure will be an important topic for further research. It is expected to meet the needs of human-like driving.

关键词： Autonomous vehicles decision-making driving behavior human-like driving

来源：评论

学校读者我要写书评

暂无评论

A Comparison of Prompt Engineering Techniques for Task Planning and Execution in Service robotics 23

A Comparison of Prompt Engineering Techniques for Task Plann...

引用

23rd IEEE-RAS International Conference on Humanoid Robots, Humanoids 2024

作者： Bode, Jonas Pätzold, Bastian Memmesheimer, Raphael Behnke, Sven Computer Science Institute Vi - Intelligent Systems and Robotics Lamarr Institute for Machine Learning and Artificial Intelligence Center for Robotics University of Bonn Autonomous Intelligent Systems group Germany

ISBN: (纸本)9798350373578

Recent advances in Large Language Models (LLMs) have been instrumental in autonomous robot control and human-robot interaction by leveraging their vast general knowledge and capabilities to understand and reason across a wide range of tasks and scenarios. Previous works have investigated various prompt engineering techniques for improving the performance of LLMs to accomplish tasks, while others have proposed methods that utilize LLMs to plan and execute tasks based on the available functionalities of a given robot platform. In this work, we consider both lines of research by comparing prompt engineering techniques and combinations thereof within the application of high-level task planning and execution in service robotics. We define a diverse set of tasks and a simple set of functionalities in simulation, and measure task completion accuracy and execution time for several state-of-the-art models. We make our code, including all prompts, available at https://***/AIS-Bonn/Prompt-Engineering. © 2024 IEEE.

关键词： Human robot interaction

来源：评论

学校读者我要写书评

暂无评论

A fast scanning strategy based on trajectory shaping for atomic force microscopy

引用

Nano Research 2022年第7期15卷 6438-6446页

作者： Yinan Wu Yingao Chang Yongchun Fang Zhi Fan Institute of Robotics and Automatic Information System College of Artificial IntelligenceNankai UniversityTianjin 300350China Tianjin Key Laboratory of Intelligent Robotics Nankai UniversityTianjin 300350China

To improve the scanning speed of an atomic force microscopy(AFM),a smooth scanning pattern is elaborately devised via trajectory shaping in this paper,so as to achieve fast imaging without hardware ***,in the proposed scanning method,the piezoelectric actuator tracks a well-designed smooth periodic signal in x-direction,and simultaneously tracks a step signal in *** advantage of the proposed method is that it does not require additional data reprocessing to construct the morphology of the sample surface,while significantly increasing the scanning bandwidth restricted by the raster scanning ***,to directly utilize the height data collected by scanning to produce the sample morphology,the forward process in the common raster scanning mode is retained in the proposed method,the tracking signal in the forward process is thus set to a ramp function in *** addition,to ensure the continuity and smoothness of the entire tracking signal in x-direction,a segment of a sine curve is uniquely determined as the backward tracking signal by position and acceleration constraints,so as to ensure that the forward and backward curves are continuous and acceleration-continuous at the intersection ***,the frequency spectrum analysis of the designed smooth signal is carried out to exhibit the depressed amplitudes of high-frequency components,which demonstrates that the proposed method is able to reduce the resonance in AFM high-speed scanning,so as to improve the capacity of rapidly generating high-quality ***,convincing comparison experiments are implemented to verify the imaging performance of the designed scanning algorithm.

关键词： atomic force microscopy fast scanning smooth signal trajectory shaping

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：