检索结果-内蒙古大学图书馆

NeSLAM: Neural Implicit Mapping and Self-Supervised Feature Tracking With Depth Completion and Denoising

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Deng, Tianchen Wang, Yanbo Xie, Hongle Wang, Hesheng Wang, Jingchuan Wang, Danwei Chen, Weidong Institute of Medical Robotics Department of Automation Shanghai Jiao Tong University Key Laboratory of System Control and Information Processing Ministry of Education Shanghai200240 China School of Electrical and Electronic Engineering Nanyang Technological University Singapore

In recent years, there have been significant advancements in 3D reconstruction and dense RGB-D SLAM systems. One notable development is the application of Neural Radiance Fields (NeRF) in these systems, which utilizes implicit neural representation to encode 3D scenes. This extension of NeRF to SLAM has shown promising results. However, the depth images obtained from consumer-grade RGB-D sensors are often sparse and noisy, which poses significant challenges for 3D reconstruction and affects the accuracy of the representation of the scene geometry. Moreover, the original hierarchical feature grid with occupancy value is inaccurate for scene geometry representation. Furthermore, the existing methods select random pixels for camera tracking, which leads to inaccurate localization and is not robust in real-world indoor environments. To this end, we present NeSLAM, an advanced framework that achieves accurate and dense depth estimation, robust camera tracking, and realistic synthesis of novel views. First, a depth completion and denoising network is designed to provide dense geometry prior and guide the neural implicit representation optimization. Second, the occupancy scene representation is replaced with Signed Distance Field (SDF) hierarchical scene representation for high-quality reconstruction and view synthesis. Furthermore, we also propose a NeRF-based self-supervised feature tracking algorithm for robust real-time tracking. Experiments on various indoor datasets demonstrate the effectiveness and accuracy of the system in reconstruction, tracking quality, and novel view synthesis. © 2024, CC BY-NC-SA.

关键词： Geometry

Supervision policies can shape long-term risk management in general-purpose AI models

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Cebrian, Manuel Gómez, Emilia Llorca, David Fernández Department of Statistics Universidad Carlos III de Madrid Spain Center for Automation and Robotics Spanish National Research Council Madrid Spain European Commission Joint Research Centre Seville Spain Computer Engineering Department University of Alcalá Alcalá de Henares Spain

The rapid proliferation and deployment of General-Purpose AI (GPAI) models, including large language models (LLMs), present unprecedented challenges for AI supervisory entities. We hypothesise that these entities will need to navigate an emergent ecosystem of risk and incident reporting, likely to exceed their supervision capacity. To investigate this, we develop a simulation framework parametrised by features extracted from the diverse landscape of risk, incident or hazard reporting ecosystems—including community-driven platforms, crowdsourcing initiatives, and expert assessments. We evaluate four supervision policies: non-prioritised (first-come, first-served), random selection, priority-based (addressing the highest-priority risks first), and diversity-prioritised (balancing high-priority risks with comprehensive coverage across risk types). Our results indicate that while priority-based and diversity-prioritised policies are more effective at mitigating high-impact risks—particularly those identified by experts—they may inadvertently neglect systemic issues reported by the broader community. This oversight can create feedback loops that amplify certain types of reporting while discouraging others, leading to a skewed perception of the overall risk landscape. We validate our simulation results with several real-world datasets, including one with over a million ChatGPT interactions, including more than 150K conversations identified as risky. This validation underscores the complex trade-offs inherent in AI risk supervision and highlights how the choice of risk management policies can shape the future landscape of AI risks across diverse GPAI models used in society. © 2025, CC BY.

关键词： Digital elevation model

Robust Control of An Aerial Manipulator Based on A Variable Inertia Parameters Model

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhang, Guangyu He, Yuqing Dai, Bo Gu, Feng Han, Jianda Liu, Guangjun The State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China Institute for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110016 China The Department of Aerospace Engineering Ryerson University TorontoONM5B 2K3 Canada

Aerial manipulator, which is composed of an UAV (Unmanned Aerial Vehicle) and a multi-link manipulator and can perform aerial manipulation, has shown great potential of applications. However, dynamic coupling between the UAV and the manipulator makes it difficult to control the aerial manipulator with high performance. In this paper, system modeling and control problem of the aerial manipulator are studied. Firstly, an UAV dynamic model is proposed with consideration of the dynamic coupling from an attached manipulator, which is treated as disturbance for the UAV. In the dynamic model, the disturbance is affected by the variable inertia parameters of the aerial manipulator system. Then, based on the proposed dynamic model, a disturbance compensation robust H∞ controller is designed to stabilize flight of the UAV while the manipulator is in operation. Finally, experiments are conducted and the experimental results demonstrate the feasibility and validity of the proposed control scheme. © 2024, CC BY.

关键词： Antennas

Research of Control Systems for Robotic Spatial Planning Platforms 4

学校读者我要写书评

暂无评论

Research of Control Systems for Robotic Spatial Planning Pla...

4th IEEE International Conference on Modern Electrical and Energy System, MEES 2022

作者： Lebedenko, Yurii Polyvoda, Oksana Derkachenko, Anatolii Modlo, Yevhenii Demishonkova, Svitlana Pylypenko, Yurii Kyiv National University of Technologies and Design Information and Computer Technologies Department Kyiv Ukraine Kherson National Technical University Automation Robotics and Mechatronics Department Kherson Ukraine Technological Institute State University of Economics and Technology Kryvyi Rih Ukraine Kyiv National University of Technologies and Design Department of Computer Engineering and Electromechanics Kyiv Ukraine

ISBN: (纸本)9798350346831

The work is devoted to the study of the control system of the moving platform of the manipulative robot. The purpose of the work is the development and research of an instrumental complex of control, analysis and modeling of the operation of positioning platforms based on a spherical mechanism for use in automated industrial systems. The state of use of multi-link positioning platforms in various industries was analyzed and their classification was provided, the main areas of use were determined. A comparative analysis of mathematical methods for describing the kinematic properties of positioning platforms was performed, namely: the closed vector contour method, the coordinate transformation method. A model with an analytical apparatus for controlling the drive links of the spherical mechanism according to the given scenarios of the orientation of the positioning platform was developed, as well as a mathematical model of inverse kinematics was made, the working space was modeled, and the service angles were determined, and a calculation and real-life experiment was conducted. For the first time, the work uses the dependencies of analytical geometry to implement algorithms for controlling a spherical mechanism by inverse kinematics, and offers software and hardware that combines visual modeling tools with real-time data transmission to a controller that controls the drivers of a spherical mechanism by direct kinematics. The result of the field experiment is the testing of the proposed control algorithm on the manufactured manipulator mechanism on the most common movements of the platform of the spherical mechanism and its exit to the extreme positions in order to determine the working space and the values of the service angles. A system of remote control of the spherical mechanism at the hardware level with an interactive interface and tracking of the positioning target has been implemented. © 2022 IEEE.

关键词： Inverse kinematics

Application of Deep Learning Method to Estimate Bottomhole Pressure Dynamics of Oil Wells

学校读者我要写书评

暂无评论

Application of Deep Learning Method to Estimate Bottomhole P...

IEEE International Symposium on Industrial Electronics (ISIE)

作者： Haibo Cheng Shichao Li Peng Zeng Valeriy Vyatkin State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang China Key Laboratory of Networked Control Systems Chinese Academy of Sciences Shenyang China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang China Department of Computer Science Electrical and Space Engineering Luleå University of Technology Luleå Sweden Department of Electrical Engineering and Automation Aalto University Helsinki Finland

Surrogate models, which have become an effective and popular method to close loop reservoir management problems, use a data-driven approach to predict dynamic injection and production wells parameters and optimize waterflooding development. In this study, a deep learning-based surrogate model method is proposed to estimate bottomhole pressure (BHP) of production wells in waterflooding reservoirs. Bidirectional long short-term memory (BiLSTM) network, as an efficient deep learning approach, is applied to BHP estimation using fluctuation data. Extended Fourier amplitude sensitivity test (EFAST) method is employed to analyse the influence of different input factors on BHP dynamics, and a reduced dataset is rebuilt to predict BHP parameter based on BiLSTM-EFAST algorithm. The estimation results are tested on a dataset from Volve oilfield in North Sea, and compared with other deep learning methods. The test results indicate that the proposed method can achieve higher prediction accuracy. A reduced dataset-based approach provides a new attempt to reduce model complexity and improve calculation speed for big data-driven surrogate model in oil and gas industry.

关键词：

Machine learning algorithm review for the detection of formalin in fish

学校读者我要写书评

暂无评论

Machine learning algorithm review for the detection of forma...

7th IET Smart Cities Symposium (SCS 2023)

作者： Adheena Benny R. Emilin Renitta X. Anitha Mary P. K. Dutta C. Karthik El-Sayed M. El-kenawy Department of Food Processing Karunya Institute of Technology and Sciences India School of Engineering and Technology Amity University Kolkata West Bengal India Department of Robotics and Automation Jyothi Engineering College Kerala India Department of Communication and Electronics Delta Higher Institute of Engineering and Technology Mansoura Egypt

Formalin is a toxic chemical that is commonly used to preserve fish and prevent spoilage, but its use in food products is illegal in many countries. The detection of Formalin in fish is critical to ensure food safety and prevent health hazards for consumers. Artificial Intelligence (AI) can be used to detect Formalin in fish samples by analyzing various parameters such as color, texture, and chemical composition. AI-based detection methods can provide fast and accurate results compared to traditional laboratory methods, which can take several hours or days. Machine learning algorithms can be trained to recognize patterns and identify the presence of Formalin in fish samples with high accuracy. AI-based detection methods can also help identify other types of adulterants and contaminants in food products, improving overall food safety. Spectroscopy techniques, such as Fourier Transform Infrared Spectroscopy (FTIR) and Raman Spectroscopy, can be combined with AI algorithms to enhance the accuracy of Formalin detection in fish. Mobilebased applications that use AI algorithms can be developed to provide on- the-spot detection of Formalin in fish samples, making it easier for food inspectors and consumers to detect adulteration. AI-based detection methods can be scaled up to handle large volumes of fish samples, making it easier to monitor the quality of fish products in the market. The use of AI-based detection methods for Formalin in fish can help promote transparency and accountability in the food industry, ensuring that consumers have access to safe and healthy food products.

关键词：

Learning Goal-oriented Bimanual Dough Rolling Using Dynamic Heterogeneous Graph Based on Human Demonstration

学校读者我要写书评

暂无评论

Learning Goal-oriented Bimanual Dough Rolling Using Dynamic ...

IEEE International Conference on robotics and Biomimetics

作者： Junjia Liu Chenzui Li Shixiong Wang Zhipeng Dong Tin Lun Lam Sylvain Calinon Miao Li Fei Chen Department of Mechanical and Automation Engineering T-Stone Robotics Institute The Chinese University of Hong Kong Hong Kong School of Science and Engineering The Chinese University of Hong Kong Shenzhen and Shenzhen Institute of Artificial Intelligence and Robotics for Society Sylvain Calinon is with the Idiap Research Institute Martigny Switzerland School of Microelectronics and the Institute of Technological Sciences Wuhan University Wuhan China

ISBN: (数字)9798331509644

ISBN: (纸本)9798331509651

Soft object manipulation poses significant chal-lenges for robots, requiring effective techniques for state representation and manipulation policy learning. State representation involves capturing the dynamic changes in the environment, while manipulation policy learning focuses on establishing the relationship between robot actions and state transformations to achieve specific goals. To address these challenges, this research paper introduces a novel approach: a dynamic heterogeneous graph-based model for learning goal-oriented soft object manipulation policies. The proposed model utilizes graphs as a unified representation for both states and policy learning. By leveraging the dynamic graph, we can extract crucial information regarding object dynamics and manipulation policies. Furthermore, the model facilitates the integration of demonstrations, enabling guided policy learning. To evaluate the efficacy of our approach, we designed a dough rolling task and conducted experiments using both a differentiable simulator and a real-world humanoid robot. Additionally, several ablation studies were performed to analyze the effect of our method, demonstrating its superiority in achieving human-like behavior.

关键词： Robot kinematics Biomimetics Pipelines Humanoid robots Interference Robustness Pins Data mining Manipulator dynamics

Automated engineered-stone silicosis screening and staging using Deep Learning with X-rays

学校读者我要写书评

暂无评论

Computers in Biology and Medicine 2025年 191卷 110153-110153页

作者： Priego-Torres, Blanca Sanchez-Morillo, Daniel Khalili, Ebrahim Conde-Sánchez, Miguel Ángel García-Gámez, Andrés León-Jiménez, Antonio Bioengineering Automation and Robotics Research Group Department of Automation Engineering Electronics and Computer Architecture and Networks School of Engineering University of Cadiz Cádiz Puerto Real11519 Spain Puerta del Mar University Hospital Cádiz11009 Spain Pulmonology Department Puerta del Mar University Hospital Cádiz11009 Spain Radiology Department Puerta del Mar University Hospital Cádiz11009 Spain Radiology Department Puerto Real University Hospital Cádiz Puerto Real11510 Spain

Silicosis, a debilitating occupational lung disease caused by inhaling crystalline silica, continues to be a significant global health issue, especially with the increasing use of engineered stone (ES) surfaces containing high silica content. Traditional diagnostic methods, dependent on radiological interpretation, have low sensitivity, especially, in the early stages of the disease, and present variability between evaluators. This study explores the efficacy of deep learning techniques in automating the screening and staging of silicosis using chest X-ray images. Utilizing a comprehensive dataset, obtained from the medical records of a cohort of workers exposed to artificial quartz conglomerates, we implemented a preprocessing stage for rib-cage segmentation, followed by classification using state-of-the-art deep learning models. The segmentation model exhibited high precision, ensuring accurate identification of thoracic structures. In the screening phase, our models achieved near-perfect accuracy, with ROC AUC values reaching 1.0, effectively distinguishing between healthy individuals and those with silicosis. The models demonstrated remarkable precision in the staging of the disease. Nevertheless, differentiating between simple silicosis and progressive massive fibrosis, the evolved and complicated form of the disease, presented certain difficulties, especially during the transitional period, when assessment can be significantly subjective. Notwithstanding these difficulties, the models achieved an accuracy of around 81% and ROC AUC scores nearing 0.93. This study highlights the potential of deep learning to generate clinical decision support tools to increase the accuracy and effectiveness in the diagnosis and staging of silicosis, whose early detection would allow the patient to be moved away from all sources of occupational exposure, therefore constituting a substantial advancement in occupational health diagnostics. © 2025 The Authors

关键词： Occupational diseases

Incremental Joint Learning of Depth, Pose and Implicit Scene Representation on Monocular Camera in Large-scale Scenes

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Deng, Tianchen Wang, Nailin Wang, Chongdi Yuan, Shenghai Wang, Jingchuan Wang, Danwei Chen, Weidong The Institute of Medical Robotics Department of Automation Shanghai Jiao Tong University Key Laboratory of System Control and Information Processing Ministry of Education Shanghai200240 China The School of Electrical and Electronic Engineering Nanyang Technological University Singapore

Dense scene reconstruction for photo-realistic view synthesis has various applications, such as VR/AR, autonomous vehicles. However, most existing methods have difficulties in large-scale scenes due to three core challenges: (a) inaccurate depth input. Accurate depth input is impossible to get in real-world large-scale scenes. (b) inaccurate pose estimation. Existing methods are not robust enough with the growth of cumulative errors in large scenes and long sequence. (c) insufficient scene representation capability. A single global radiance field lacks the capacity to effectively scale to large-scale scenes. To this end, we propose an incremental joint learning framework, which can achieve accurate depth, pose estimation, and large-scale scene reconstruction. A vision transformer-based network is adopted as the backbone to enhance performance in scale information estimation. For pose estimation, a feature-metric bundle adjustment (FBA) method is designed for accurate and robust camera tracking in large-scale scenes. In terms of implicit scene representation, we propose an incremental scene representation method to construct the entire large-scale scene as multiple local radiance fields to enhance the scalability of 3D scene representation. Extended experiments have been conducted to demonstrate the effectiveness and accuracy of our method in depth estimation, pose estimation, and large-scale scene reconstruction. The code will be open-sourced upon paper acceptance. © 2024, CC BY-NC-SA.

关键词： Depth perception