检索结果-内蒙古大学图书馆

RelMobNet: End-to-end relative camera pose estimation using a robust two-stage training

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Rajendran, Praveen Kumar Mishra, Sumit Vecchietti, Luiz Felipe Har, Dongsoo Division of Future Vehicle KAIST Korea Republic of The Robotics Program KAIST Korea Republic of Data Science Group Institute for Basic Science Korea Republic of The CCS Graduate School of Mobility KAIST Korea Republic of

Relative camera pose estimation, i.e. estimating the translation and rotation vectors using a pair of images taken in different locations, is an important part of systems in augmented reality and robotics. In this paper, we present an end-to-end relative camera pose estimation network using a siamese architecture that is independent of camera parameters. The network is trained using the Cambridge Landmarks data with four individual scene datasets and a dataset combining the four scenes. To improve generalization, we propose a novel two-stage training that alleviates the need of a hyperparameter to balance the translation and rotation loss scale. The proposed method is compared with one-stage training CNN-based methods such as RPNet and RCPNet and demonstrate that the proposed model improves translation vector estimation by 16.11%, 28.88%, and 52.27% on the Kings College, Old Hospital, and St Marys Church scenes, respectively. For proving texture invariance, we investigate the generalization of the proposed method augmenting the datasets to different scene styles, as ablation studies, using generative adversarial networks. Also, we present a qualitative assessment of epipolar lines of our network predictions and ground truth poses. © 2022, CC BY.

关键词： Cameras

Reward Generation via Large Vision-Language Model in Offline Reinforcement Learning

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Lee, Younghwan Luu, Tung M. Lee, Donghoon Yoo, Chang D. Electrical Engineering KAIST Daejeon Korea Republic of Robotics Program KAIST Daejeon Korea Republic of

In offline reinforcement learning (RL), learning from fixed datasets presents a promising solution for domains where real-time interaction with the environment is expensive or risky. However, designing dense reward signals for offline dataset requires significant human effort and domain expertise. Reinforcement learning with human feedback (RLHF) has emerged as an alternative, but it remains costly due to the human-in-the-loop process, prompting interest in automated reward generation models. To address this, we propose Reward Generation via Large Vision-Language Models (RG-VLM), which leverages the reasoning capabilities of LVLMs to generate rewards from offline data without human involvement. RG-VLM improves generalization in long-horizon tasks and can be seamlessly integrated with the sparse reward signals to enhance task performance, demonstrating its potential as an auxiliary reward signal. © 2025, CC BY.

关键词： Reinforcement learning

Development of Cleaning Module for Wall Climbing Drone with Bio-inspired Watering Mechanism

学校读者我要写书评

暂无评论

Development of Cleaning Module for Wall Climbing Drone with ...

International Conference on Control, Automation and Systems ( ICCAS)

作者： Taewan Kim Byeongho Yu Christian Tirtawardhana I Made Aswin Nahrendra Hyun Myung Robotics Program Korea Advanced Institute of Science and Technology (KAIST) Daejeon South Korea School of Electrical Engineering Korea Advanced Institute of Science and Technology (KAIST) Daejeon South Korea

For many years, there has been significant progress in cleaning facilities in hazardous conditions using autonomous mechanical systems to keep humans away from potential danger. Especially, the cleaning method using a UAV (Unmanned Aerial Vehicle) has attracted much interest in the robotics field due to its high degree of freedom in maneuverability; however, a UAV has a payload limitation. In this paper, we propose a cleaning module for a specific drone platform “CAROS-Q” that can tilt its pose to perch on walls or tilted surfaces. The proposed module includes bio-inspired water flow control mechanism from the peristaltic movement of the Esophagus for improving cleaning performance while minimizing the complexity and weight. By comparing the cleaning performance between the proposed module and dry brush cleaning, it was confirmed that the proposed module outperforms the counterpart for cleaning various contaminants.

关键词： Brushes Surface contamination Ink Glass Autonomous aerial vehicles Cleaning Robots

Learning Terminal State of the Trajectory Planner: Application for Collision Scenarios of Autonomous Vehicles

学校读者我要写书评

暂无评论

Learning Terminal State of the Trajectory Planner: Applicati...

IEEE International Conference on robotics and Automation (ICRA)

作者： Joonhee Lim Kibeom Lee Jangho Shin Dongsuk Kum Robotics Program Korea Advanced Institute of Science and Technology (KAIST) Daejeon Republic of Korea Department of Mechanical Engineering Gachon University Seongnam Republic of Korea Research & Development Division Hyundai Motor Company Hwaseong Republic of Korea The Cho Chun Shik Graduate School of Mobility Korea Advanced Institute of Science and Technology (KAIST) Daejeon Republic of Korea

ISBN: (数字)9798350384574

ISBN: (纸本)9798350384581

Collision Avoidance/Mitigation System (CAMS) for autonomous vehicles is a crucial technology that ensures the safety and reliability of autonomous driving systems. Conventional collision avoidance approaches struggle in complex and various scenarios by avoiding collisions based on rules for specific collision scenarios. This has led to learning-based methods using neural networks for adaptive collision avoidance. However, the approaches directly outputting control inputs through neural networks have drawbacks in interpretability and stability. To address these limitations, we propose a trajectory planning method for CAMS that combines deep reinforcement learning (DRL) and quintic polynomial (QP) trajectory planning. The proposed method determines the terminal state and confidence of the trajectory using DRL and plans a QP trajectory based on them. By utilizing the terminal state and confidence of the trajectory rather than direct control inputs as the output of the neural network, it generates a more realistic and continuous path. Moreover, this approach considers collision avoidance and mitigation in an integrated manner through the reward function of RL. Our experimental results demonstrate that the proposed method not only improves interpretability and stability compared to existing learning-based methods but also upholds performance in complex and various collision scenarios.

关键词： Learning systems Trajectory planning Neural networks Trajectory Cams Safety Reliability

Ambiguity-Aware Multi-Object Pose Optimization for Visually-Assisted Robot Manipulation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Jeon, Myung-Hwan Kim, Jeongyun Ryu, Jee-Hwan Kim, Ayoung Dept. of the Robotics Program KAIST Daejeon Korea Republic of Dept. of Mechanical Engineering SNU Seoul Korea Republic of Dept. of Civil and Environmental Engineering KAIST Daejeon Korea Republic of

—6D object pose estimation aims to infer the relative pose between the object and the camera using a single image or multiple images. Most works have focused on predicting the object pose without associated uncertainty under occlusion and structural ambiguity (symmetricity). However, these works demand prior information about shape attributes, and this condition is hardly satisfied in reality;even asymmetric objects may be symmetric under the viewpoint change. In addition, acquiring and fusing diverse sensor data is challenging when extending them to robotics applications. Tackling these limitations, we present an ambiguity-aware 6D object pose estimation network, PrimA6D++, as a generic uncertainty prediction method. The major challenges in pose estimation, such as occlusion and symmetry, can be handled in a generic manner based on the measured ambiguity of the prediction. Specifically, we devise a network to reconstruct the three rotation axis primitive images of a target object and predict the underlying uncertainty along each primitive axis. Leveraging the estimated uncertainty, we then optimize multi-object poses using visual measurements and camera poses by treating it as an object SLAM problem. The proposed method shows a significant performance improvement in T-LESS and YCB-Video datasets. We further demonstrate real-time scene recognition capability for visually-assisted robot manipulation. Our code and supplementary materials are available at https://***/rpmsnu/PrimA6D. Copyright © 2022, The Authors. All rights reserved.

关键词： Cameras

AllWeather-UsableWearable Dual Energy Harvester for Outdoor Sustainable Operation

学校读者我要写书评

暂无评论

SusMat 2025年第1期5卷 159-170页

作者： Yeongju Jung Jiyong Ahn Ji-Seok Kim Junhyuk Bang Minwoo Kim Seongmin Jeong Jinwoo Lee Seonggeun Han Il-Kwon Oh Seung Hwan Ko Applied Nano and Thermal Science Lab Department of Mechanical EngineeringSeoul National UniversitySeoulRepublic of Korea National Creative Research Initiative for Functionally Antagonistic Nano-Engineering Department of Mechanical EngineeringSchool of Mechanical and Aerospace EngineeringKorea Advanced Institute of Science and Technology(KAIST)DaejeonRepublic of Korea Department of Mechanical Roboticsand Energy EngineeringDongguk UniversitySeoulRepublic of Korea Device Solutions Research Samsung ElectronicsHwaseong-siRepublic of Korea Institute of Engineering Research/Institute of Advanced Machinery and Design(SNU-IAMD) Seoul National UniversitySeoulRepublic of Korea Interdisciplinary Program in Bioengineering Seoul National UniversitySeoulRepublic of Korea

Sustainable power sources for outdoor wearable electronics are essential for the continuous operation of wearable ***,the current lack of engineering design that can harvest energy regardless of weather conditions presents a significant *** this regard,this study introduces a wearable,breathable all-weather usable dual energy harvester(AWuDEH)that can seamlessly generate electrical energy regardless of weather *** this study,the AWuDEH integrated with the thermoelectric generator and the droplet-based electricity generator is *** AWuDEH,especially engineered with a bi-functional top substrate for radiative cooling and electrification,achieves sustainable energy harvesting outdoors,thereby addressing the conventional challenge associated with the necessity for separate energy harvesters tailored to outdoor usage contingent on weather *** device reaches a maximum power output of 14.6μW cm^(-2)under simulated sunny conditions and generates a much more enhanced thermoelectric power of 74.78μW cm^(-2)and a droplet-based electric power of 256.25 mW m^(-2)in rainy *** proof,this study developed self-powered wearable electronics capable of acquiring physiological signals in simulated outdoor *** study presents a promising advancement in wearable technology,offering a potent solution for sustainable energy harvesting independent of weather conditions.

关键词： all-weather usable droplet-based electricity generator hybrid architecture outdoor sustainable operation thermoelectric generator

SC-LiDAR-SLAM: a Front-end Agnostic Versatile LiDAR SLAM System

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Kim, Giseop Yun, Seungsang Kim, Jeongyun Kim, Ayoung Dept. of Civil and Envtl. Eng. KAIST Daejeon Korea Republic of Robotics Program KAIST Daejeon Korea Republic of Dept. of Mechanical Eng Seoul National University Seoul Korea Republic of

Accurate 3D point cloud map generation is a core task for various robot missions or even for data-driven urban analysis. To do so, light detection and ranging (LiDAR) sensor-based simultaneous localization and mapping (SLAM) technology have been elaborated. To compose a full SLAM system, many odometry and place recognition methods have independently been proposed in academia. However, they have hardly been integrated or too tightly combined so that exchanging (upgrading) either single odometry or place recognition module is very effort demanding. Recently, the performance of each module has been improved a lot, so it is necessary to build a SLAM system that can effectively integrate them and easily replace them with the latest one. In this paper, we release such a front-end agnostic LiDAR SLAM system, named SC-LiDAR-SLAM. We built a complete SLAM system by designing it modular, and successfully integrating it with Scan Context++ and diverse existing open-source LiDAR odometry methods to generate an accurate point cloud map. © 2022, CC BY.

关键词： Optical radar

ERASOR: Egocentric ratio of pseudo occupancy-based dynamic object removal for static 3D point cloud map building

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Lim, Hyungtae Hwang, Sungwon Myung, Hyun Daejeon34141 Korea Republic of Robotics Program at KAIST Daejeon34141 Korea Republic of

Scan data of urban environments often include representations of dynamic objects, such as vehicles, pedestrians, and so forth. However, when it comes to constructing a 3D point cloud map with sequential accumulations of the scan data, the dynamic objects often leave unwanted traces in the map. These traces of dynamic objects act as obstacles and thus impede mobile vehicles from achieving good localization and navigation performances. To tackle the problem, this paper presents a novel static map building method called ERASOR, Egocentric RAtio of pSeudo Occupancy-based dynamic object Removal, which is fast and robust to motion ambiguity. Our approach directs its attention to the nature of most dynamic objects in urban environments being inevitably in contact with the ground. Accordingly, we propose the novel concept called pseudo occupancy to express the occupancy of unit space and then discriminate spaces of varying occupancy. Finally, Region-wise Ground Plane Fitting (R-GPF) is adopted to distinguish static points from dynamic points within the candidate bins that potentially contain dynamic points. As experimentally verified on SemanticKITTI, our proposed method yields promising performance against state-of-the-art methods overcoming the limitations of existing ray tracing-based and visibility-based methods. Copyright © 2021, The Authors. All rights reserved.

关键词： Urban planning

Multi-session Underwater Pose-graph SLAM using Inter-session Opti-acoustic Two-view Factor

学校读者我要写书评

暂无评论

Multi-session Underwater Pose-graph SLAM using Inter-session...

IEEE International Conference on robotics and Automation (ICRA)

作者： Hyesu Jang Sungho Yoon Ayoung Kim Depart. of Civil and Env. Eng KAIST Daejeon S. Korea Robotics Program KAIST Daejeon S. Korea

Concurrent mapping necessitates data association among vehicles to overcome temporal and sensor modality differences. In this work, we focus on an underwater multi-vehicle mapping scenario in which vehicles have various sensor modalities, namely sonar and camera. This inter-session sonar-optical image matching poses two main challenges. First, ensuring covisibility for the opti-acoustic pair is complex due to their projection models and field of view (FOV) difference. Second, even with secured covisible frames, feature matching over various sensor modalities is not trivial. To overcome these challenges, we complete multi-session simultaneous localization and mapping (SLAM) by introducing an opti-acoustic pairwise factor. We alleviate the covisibility requirement by introducing inter-session measurements. We achieved opti-acoustic feature matching by applying a style-transfer and integration with SuperGlue. The proposed method is validated via simulation and real underwater tank tests.

关键词： Integrated optics Simultaneous localization and mapping Automation Image matching Conferences Sonar Optical imaging