检索结果-内蒙古大学图书馆

deep Weakly Supervised Positioning for Indoor Mobile Robots

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第2期7卷 1206-1213页

作者： Wang, Ruoyu Xu, Xuchu Ding, Li Huang, Yang Feng, Chen NYU Brooklyn NY 11201 USA Univ Rochester Rochester NY 14627 USA Univ Calif Berkeley Berkeley CA 94720 USA

PoseNet can map a photo to the position where it is taken, which is appealing in robotics. However training PoseNet requires full supervision, where ground truth positions are non-trivial to obtain. Can we train PoseNet without knowing the ground truth positions for each observation? We show that it is possible to do so via constraint-based weak-supervision, leading to the proposed framework: deepGPS. Particularly, using wheel-encoder-estimated distances traveled by a robot along with random straight line segments as constraints between PoseNet outputs, deepGPS can achieve a relative positioning error of less than 2% for indoor robot positioning. Moreover, training deepGPS can be done as auto-calibration with almost no human attendance, which is more attractive than its competing methods that typically require careful and expert-level manual calibration. We conduct various experiments on simulated and real datasets to demonstrate the general applicability, effectiveness, and accuracy of deepGPS on indoor mobile robots and perform a comprehensive analysis of its robustness.

关键词： deep learning for visual perception localization

来源：评论

学校读者我要写书评

暂无评论

MultiRes-NetVLAD: Augmenting Place Recognition Training With Low-Resolution Imagery

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第2期7卷 3882-3889页

作者： Khaliq, Ahmad Milford, Michael Garg, Sourav Queensland Univ Technol Sch Elect Engn & Robot Brisbane Qld 4000 Australia Queensland Univ Technol QUT Ctr Robot Brisbane Qld 4000 Australia

visual Place Recognition (VPR) is a crucial component of 6-DoF localization, visual SLAM and structure-from-motion pipelines, tasked to generate an initial list of place match hypotheses by matching global place descriptors. However, commonly-used CNN-based methods either process multiple image resolutions after training or use a single resolution and limit multi-scale feature extraction to the last convolutional layer during training. In this paper, we augment NetVLAD representation learning with low-resolution image pyramid encoding which leads to richer place representations. The resultant multi-resolution feature pyramid can be conveniently aggregated through VLAD into a single compact representation, avoiding the need for concatenation or summation of multiple patches in recent multi-scale approaches. Furthermore, we show that the underlying learnt feature tensor can be combined with existing multi-scale approaches to improve their baseline performance. Evaluation on 15 viewpoint-varying and viewpoint-consistent benchmarking datasets confirm that the proposed MultiRes-NetVLAD leads to state-of-the-art Recall@N performance for global descriptor based retrieval, compared against 11 existing techniques. Source code is publicly available at https://***/Ahmedest61/MultiRes-NetVLAD.

关键词： Image resolution Training visualization Representation learning Image recognition Feature extraction Task analysis Localization deep learning for visual perception representation learning visual place recognition multi-scale

来源：评论

学校读者我要写书评

暂无评论

Particle Filters in Latent Space for Robust Deformable Linear Object Tracking

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第4期7卷 12577-12584页

作者： Yang, Yuxuan Stork, Johannes A. A. Stoyanov, Todor Orebro Univ Ctr Appl Autonomous Sensor Syst AASS Autonomous Mobile Manipulat Lab Orebro 70182 Sweden

Tracking of deformable linear objects (DLOs) is important for many robotic applications. However, achieving robust and accurate tracking is challenging due to the lack of distinctive features or appearance on the DLO, the object's high-dimensional state space, and the presence of occlusion. In this letter, we propose a method for tracking the state of a DLO by applying a particle filter approach within a lower-dimensional state embedding learned by an autoencoder. The dimensionality reduction preserves state variation, while simultaneously enabling a particle filter to accurately track DLO state evolution with a practically feasible number of particles. Compared to previous works, our method requires neither running a high-fidelity physics simulation, nor manual designs of constraints and regularization. Without the assumption of knowing the initial DLO state, our method can achieve accurate tracking even under complex DLO motions and in the presence of severe occlusions.

关键词： deep learning for visual perception perception for grasping and manipulation RGB-D perception

来源：评论

学校读者我要写书评

暂无评论

3D Part Assembly Generation With Instance Encoded Transformer

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第4期7卷 9051-9058页

作者： Zhang, Rufeng Kong, Tao Wang, Weihao Han, Xuan You, Mingyu Tongji Univ Coll Elect & Informat Engn Frontiers Sci Ctr Intelligent Autonomous Syst Shanghai 201800 Peoples R China ByteDance AI Lab Beijing 100089 Peoples R China

It is desirable to enable robots capable of automatic assembly. Structural understanding of object parts plays a crucial role in this task yet remains relatively unexplored. In this letter, we focus on the setting of furniture assembly from a complete set of part geometries, which is essentially a 6-DoF part pose estimation problem. We propose a multi-layer transformer-based framework that involves geometric and relational reasoning between parts to update the part poses iteratively. We carefully design a unique instance encoding to solve the ambiguity between geometrically-similar parts so that all parts can be distinguished. In addition to assembling from scratch, we extend our framework to a new task called in-process part assembly. Analogous to furniture maintenance, it requires robots to continue with unfinished products and assemble the remaining parts into appropriate positions. Our method achieves far more than 10% improvements over the current state-of-the-art in multiple metrics on the public PartNet dataset. Extensive experiments and quantitative comparisons demonstrate the effectiveness of the proposed framework.

关键词： Assembly deep learning for visual perception AI-Enabled robotics

来源：评论

学校读者我要写书评

暂无评论

Contrastive 3D Shape Completion and Reconstruction for Agricultural Robots Using RGB-D Frames

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第4期7卷 10120-10127页

作者： Magistri, Federico Marks, Elias Nagulavancha, Sumanth Vizzo, Ignacio Laeebe, Thomas Behley, Jens Halstead, Michael McCool, Chris Stachniss, Cyrill Univ Bonn D-53111 Bonn Germany Univ Oxford Dept Engn Sci Oxford OX1 2JD England Lamas Inst Machine Learning & Artificial Intellig D-53757 St Augustin Germany

Monitoring plants and fruits is important in modern agriculture, with applications ranging from high-throughput phenotyping to autonomous harvesting. Obtaining highly accurate 3D measurements under real agricultural conditions is a challenging task. In this letter, we address the problem of estimating the 3D shape of fruits when only a partial view is available. We propose a pipeline that exploits high-resolution 3D data in the learning phase but only requires a single RGB-D frame to predict the 3D shape of a complete fruit during operation. To achieve this, we first learn a latent space of potential fruit appearances that we can decode into an SDF volume. With the pretrained, frozen decoder, we subsequently learn an encoder that can produce meaningful latent vectors from a single RGB-D frame. The experiments presented in this letter suggest that our approach can predict the 3D shape of whole fruits online, needing only 4 ms for inference. We evaluate our approach in controlled environments and illustrate its deployment in greenhouses without modifications.

关键词： deep learning for visual perception robotics and automation in agriculture and forestry RGB-D perception

来源：评论

学校读者我要写书评

暂无评论

A Sim-to-Real Object Recognition and Localization Framework for Industrial Robotic Bin Picking

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第2期7卷 3961-3968页

作者： Li, Xianzhi Cao, Rui Feng, Yidan Chen, Kai Yang, Biqi Fu, Chi-Wing Li, Yichuan Dou, Qi Liu, Yun-Hui Heng, Pheng-Ann Chinese Univ Hong Kong Dept Comp Sci & Engn Hong Kong Peoples R China Chinese Univ Hong Kong Dept Mech & Automat Engn Hong Kong Peoples R China

We present a generic and robust sim-to-real deep-learning-based framework, namely S2R-Pick, for fast and accurate object recognition and localization in industrial robotic bin picking. Unlike existing works designed for general everyday environments, objects for industrial bin picking are often texture-less, metallic, and also suffer from severe occlusion and cluttering. In this work, we first develop an automated synthetic data generation pipeline to produce large-scale photo-realistic data with precise annotations. We then formulate an instance segmentation network for object recognition and a 6D pose estimation network for localization. We cascade and train the two networks only on our generated synthetic data and can directly transfer them for processing real inputs without needing to retrain them on real samples. Extensive experiments show the effectiveness and superiority of our framework over state-of-the-art methods.

关键词： Pose estimation Image segmentation Three-dimensional displays Robots Object recognition Gray-scale Solid modeling deep learning for visual perception industrial robots bin picking

来源：评论

学校读者我要写书评

暂无评论

Neural Scene Representation for Locomotion on Structured Terrain

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第4期7卷 8667-8674页

作者： Hoeller, David Rudin, Nikita Choy, Christopher Anandkumar, Animashree Hutter, Marco Swiss Fed Inst Technol Robot Syst Lab CH-8050 Zurich Switzerland NVIDIA Santa Clara CA 94305 USA CALTECH Pasadena CA 91125 USA

We propose a learning-based method to reconstruct the local terrain for locomotion with a mobile robot traversing urban environments. Using a stream of depth measurements from the onboard cameras and the robot's trajectory, the algorithm estimates the topography in the robot's vicinity. The raw measurements from these cameras are noisy and only provide partial and occluded observations that in many cases do not show the terrain the robot stands on. Therefore, we propose a 3D reconstruction model that faithfully reconstructs the scene, despite the noisy measurements and large amounts of missing data coming from the blind spots of the camera arrangement. The model consists of a 4D fully convolutional network on point clouds that learns the geometric priors to complete the scene from the context and an auto-regressive feedback to leverage spatio-temporal consistency and use evidence from the past. The network can be solely trained with synthetic data, and due to extensive augmentation, it is robust in the real world, as shown in the validation on a quadrupedal robot, ANYmal, traversing challenging settings. We run the pipeline on the robot's onboard low-power computer using an efficient sparse tensor implementation and show that the proposed method outperforms classical map representations.

关键词： Representation learning deep learning for visual perception

来源：评论

学校读者我要写书评

暂无评论

Homography-Based Loss Function for Camera Pose Regression

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第3期7卷 6242-6249页

作者： Boittiaux, Clementin Marxer, Ricard Dune, Claire Arnaubec, Aurelien Hugel, Vincent IFREMER F-83000 La Seyne Sur Mer France Aix Marseille Univ Univ Toulon CNRS LIS F-83000 Toulon France Univ Toulon & Var COSMER F-83000 Toulon France

Some recent visual-based relocalization algorithms rely on deep learning methods to perform camera pose regression from image data. This letter focuses on the loss functions that embed the error between two poses to perform deep learning based camera pose regression. Existing loss functions are either difficult-to-tune multi-objective functions or present unstable reprojection errors that rely on ground truth 3D scene points and require a two-step training. To deal with these issues, we introduce a novel loss function which is based on a multiplane homography integration. This new function does not require prior initialization and only depends on physically interpretable hyperparameters. Furthermore, the experiments carried out on well established relocalization datasets show that it minimizes best the mean square reprojection error during training when compared with existing loss functions.

关键词： Localization deep learning for visual perception

来源：评论

学校读者我要写书评

暂无评论

Self-Supervised Scene Flow Estimation With 4-D Automotive Radar

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第3期7卷 8233-8240页

作者： Ding, Fangqiang Pan, Zhijun Deng, Yimin Deng, Jianning Lu, Chris Xiaoxuan Univ Edinburgh Sch Informat Edinburgh EH8 9AF Midlothian Scotland SAIC Motor Tech Ctr Shanghai 200041 Peoples R China

Scene flow allows autonomous vehicles to reason about the arbitrary motion of multiple independent objects which is the key to long-term mobile autonomy. While estimating the scene flow from LiDAR has progressed recently, it remains largely unknown how to estimate the scene flow from a 4-D radar - an increasingly popular automotive sensor for its robustness against adverse weather and lighting conditions. Compared with the LiDAR point clouds, radar data are drastically sparser, noisier and in much lower resolution. Annotated datasets for radar scene flow are also in absence and costly to acquire in the real world. These factors jointly pose the radar scene flow estimation as a challenging problem. This work aims to address the above challenges and estimate scene flow from 4-D radar point clouds by leveraging self-supervised learning. A robust scene flow estimation architecture and three novel losses are bespoken designed to cope with intractable radar data. Real-world experimental results validate that our method is able to robustly estimate the radar scene flow in the wild and effectively supports the downstream task of motion segmentation.

关键词： deep learning for visual perception visual learning automotive radars scene flow estimation

来源：评论

学校读者我要写书评

暂无评论

Recommending Fine-Grained Tool Consistent With Common Sense Knowledge for Robot

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2022年第4期7卷 8574-8581页

作者： Xin, Jianjia Wang, Lichun Wang, Shaofan Liu, Yukun Yang, Chao Yin, Baocai Beijing Univ Technol Fac Informat Technol Beijing Artificial Intelligence Inst Beijing Key Lab Multimedia & Intelligent Software Beijing 100124 Peoples R China

When robots carry out task, selecting an appropriate tool is necessary. The current research ignores the fine-grained characteristic of tasks, and mainly focuses on whether the task can be completed. Little consideration is paid for the object being manipulated, which affects the task completion quality. In order to support task oriented fine-grained tool recommendation, based on common sense knowledge, this paper proposes Fine-grained Tool-Task Graph (FTTG) to describe multi-granularity semantics of tasks, tools, objects being manipulated and relationships among them. According to FTTG, a Fine-grained Tool-Task (FTT) dataset is constructed by labeling images of tools and objects being manipulated with the defined semantics. A baseline method named Fine-grained Tool Recommendation Network (FTR-Net) is also proposed in this paper. FTR-Net gives coarse-grained and fine-grained semantic predictions by simultaneously learning the common and special features of the tools and objects being manipulated. At the same time, FTR-Net constrains the distance between features of the well matched tool and object more smaller than that of those unmatched. The constraint and the special feature ensure FTR-Net provide fine-grained tool recommendation. The constraint and the common feature ensure FTR-Net provide coarse-grained tool recommendation when the fine-grained tools are not available. Experiments show that FTR-Net can recommend tools consistent with common sense whether on test data sets or in real situations.

关键词： Computer vision for automation data sets for robotic vision deep learning for visual perception

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：