检索结果-内蒙古大学图书馆

IEEE International Conference on Automation Science and Engineering (CASE)

作者： Philipp Quentin Dino Knoll Daniel Goehring BMW Group Munich Germany Freie Universitaet Berlin Dahlem Center for Machine Learning and Robotics Berlin Germany

Despite the advances in robotics a large proportion of the of parts handling tasks in the automotive industry's internal logistics are not automated but still performed by humans. A key component to competitively automate these processes is a 6D pose estimation that can handle a large number of different parts, is adaptable to new parts with little manual effort, and is sufficiently accurate and robust with respect to industry requirements. In this context, the question arises as to the current status quo with respect to these measures. To address this we built a representative 6D pose estimation pipeline with state-of-the-art components from economically scalable real to synthetic data generation to pose estimators and evaluated it on automotive parts with regards to a realistic sequencing process. We found that using the data generation approaches, the performance of the trained 6D pose estimators are promising, but do not meet industry requirements. We reveal that the reason for this is the inability of the estimators to provide reliable uncertainties for their poses, rather than the ability of to provide sufficiently accurate poses. In this context we further analyzed how RGB- and RGB-D-based approaches compare against this background and show that they are differently vulnerable to the domain gap induced by synthetic data.

关键词：

来源：评论

学校读者我要写书评

暂无评论

TReR: A Lightweight Transformer Re-Ranking Approach for 3D LiDAR Place Recognition

TReR: A Lightweight Transformer Re-Ranking Approach for 3D L...

引用

International Conference on Intelligent Transportation

作者： Tiago Barros Luís Garrote Martin Aleksandrov Cristiano Premebida Urbano J. Nunes Department of Electrical and Computer Engineering University of Coimbra Institute of Systems and Robotics Portugal Dahlem Center for Machine Learning and Robotics Freie Universität Berlin Berlin

Autonomous driving systems often require reliable loop closure detection to guarantee reduced localization drift. Recently, 3D LiDAR-based localization methods have used retrieval-based place recognition to find revisited places efficiently. However, when deployed in challenging real-world scenarios, the place recognition models become more complex, which comes at the cost of high computational demand. This work tackles this problem from an information-retrieval perspective, adopting a first-retrieve-then-re-ranking paradigm, where an initial loop candidate ranking, generated from a 3D place recognition model, is re-ordered by a proposed lightweight transformer-based re-ranking approach (TReR). The proposed approach relies on global descriptors only, being agnostic to the place recognition model. The experimental evaluation, conducted on the KITTI Odometry dataset, where we compared TReR with s.o.t.a. re-ranking approaches such as αQE and SGV, indicate the robustness and efficiency when compared to αQE while offering a good trade-off between robustness and efficiency when compared to SGV.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fuel Efficiency of Garbage Truck Navigation

Fuel Efficiency of Garbage Truck Navigation

引用

Artificial Intelligence and Internet of Things (GCAIoT), IEEE Global Conference

作者： Martin D. Aleksandrov Khaled Alomari Dahlem Center for Machine Learning and Robotics Freie Universität Berlin Berlin Germany

ISBN: (纸本)9798350309850

We investigate achieving fuel efficiency in an economy of garbage truck collections. Fuel efficiency relates to minimising both fuel consumption and travel time. Minimising just the fuel consumption is often done by garbage collection companies and minimising just the travel time is the most popular Google Maps heuristic. We combine these two criteria for online fuel-efficiency optimisation. We prove that this takes exponential time in some edge cases but polynomial time in many practical cases. We also design three natural heuristics, approximating fuel efficiency in the edge cases, that minimise fuel consumption, travel time, and both, respectively. We evaluate the performance of these heuristics in preliminary simulations. Finally, we discuss a plan for the possible deployment of these methods.

关键词： Navigation Companies Fuels Internet of Things Artificial intelligence Optimization

来源：评论

学校读者我要写书评

暂无评论

TReR: A Lightweight Transformer Re-Ranking Approach for 3D LiDAR Place Recognition

arXiv

引用

arXiv 2023年

作者： Barros, Tiago Garrote, Luís Aleksandrov, Martin Premebida, Cristiano Nunes, Urbano J. The University of Coimbra Institute of Systems and Robotics Department of Electrical and Computer Engineering Portugal Dahlem Center for Machine Learning and Robotics Freie Universität Berlin Berlin Germany

关键词： Efficiency

来源：评论

学校读者我要写书评

暂无评论

Multi-modal NeRF Self-Supervision for LiDAR Semantic Segmentation

Multi-modal NeRF Self-Supervision for LiDAR Semantic Segment...

引用

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

作者： Xavier Timoneda Markus Herb Fabian Duerr Daniel Goehring Fisher Yu Onboard Fusion Team at CARIAD SE Volkswagen Group Ingolstadt Germany Dahlem Center for Machine Learning and Robotics Group at Freie Universität Berlin Germany Department of Information Technology and Electrical Engineering ETH Zurich Switzerland

ISBN: (数字)9798350377705

ISBN: (纸本)9798350377712

LiDAR Semantic Segmentation is a fundamental task in autonomous driving perception consisting of associating each LiDAR point to a semantic label. Fully-supervised models have widely tackled this task, but they require labels for each scan, which either limits their domain or requires impractical amounts of expensive *** images, which are generally recorded alongside LiDAR pointclouds, can be processed by the widely available 2D foundation models, which are generic and dataset-agnostic. However, distilling knowledge from 2D data to improve LiDAR perception raises domain adaptation challenges. For example, the classical perspective projection suffers from the parallax effect produced by the position shift between both sensors at their respective capture *** propose a Semi-Supervised learning setup to leverage unlabeled LiDAR pointclouds alongside distilled knowledge from the camera images. To self-supervise our model on the unlabeled scans, we add an auxiliary NeRF head and cast rays from the camera viewpoint over the unlabeled voxel features. The NeRF head predicts densities and semantic logits at each sampled ray location which are used for rendering pixel semantics. Concurrently, we query the Segment-Anything (SAM) foundation model with the camera image to generate a set of unlabeled generic masks. We fuse the masks with the rendered pixel semantics from LiDAR to produce pseudo-labels that supervise the pixel predictions. During inference, we drop the NeRF head and run our model with only *** show the effectiveness of our approach in three public LiDAR Semantic Segmentation benchmarks: nuScenes, SemanticKITTI and ScribbleKITTI.

关键词： Laser radar Head Foundation models Semantic segmentation Semantics Benchmark testing Cameras Neural radiance field Rendering (computer graphics) Reliability

来源：评论

学校读者我要写书评

暂无评论

When explanations lie: Why many modified BP attributions fail 37

When explanations lie: Why many modified BP attributions fai...

引用

37th International Conference on machine learning, ICML 2020

作者： Sixt, Leon Granz, Maximilian Landgraf, Tim Dahlem Center of Machine Learning and Robotics Freie Universität Berlin Germany

ISBN: (纸本)9781713821120

Attribution methods aim to explain a neural network's prediction by highlighting the most relevant image areas. A popular approach is to backpropagate (BP) a custom relevance score using modified rules, rather than the gradient. We analyze an extensive set of modified BP methods: Deep Taylor Decomposition, Layer-wise Relevance Propagation (LRP), Excitation BP, PatternAttribution, DeepLIFT, Deconv, RectGrad, and Guided BP. We find empirically that the explanations of all mentioned methods, except for DeepLIFT, are independent of the parameters of later layers. We provide theoretical insights for this surprising behavior and also analyze why DeepLIFT does not suffer from this limitation. Empirically, we measure how information of later layers is ignored by using our new metric, cosine similarity convergence (CSC). The paper provides a framework to assess the faithfulness of new and existing modified BP methods theoretically and empirically. © 2020 37th International Conference on machine learning, ICML 2020. All rights reserved.

关键词： Backpropagation

来源：评论

学校读者我要写书评

暂无评论

Multi-modal NeRF Self-Supervision for LiDAR Semantic Segmentation

arXiv

引用

arXiv 2024年

作者： Timoneda, Xavier Herb, Markus Duerr, Fabian Goehring, Daniel Yu, Fisher Onboard Fusion team at CARIAD SE Volkswagen Group Ingolstadt Germany Dahlem Center for Machine Learning and Robotics group Freie Universität Berlin Germany Department of Information Technology and Electrical Engineering ETH Zürich Switzerland

LiDAR Semantic Segmentation is a fundamental task in autonomous driving perception consisting of associating each LiDAR point to a semantic label. Fully-supervised models have widely tackled this task, but they require labels for each scan, which either limits their domain or requires impractical amounts of expensive annotations. Camera images, which are generally recorded alongside LiDAR pointclouds, can be processed by the widely available 2D foundation models, which are generic and dataset-agnostic. However, distilling knowledge from 2D data to improve LiDAR perception raises domain adaptation challenges. For example, the classical perspective projection suffers from the parallax effect produced by the position shift between both sensors at their respective capture times. We propose a Semi-Supervised learning setup to leverage unlabeled LiDAR pointclouds alongside distilled knowledge from the camera images. To self-supervise our model on the unlabeled scans, we add an auxiliary NeRF head and cast rays from the camera viewpoint over the unlabeled voxel features. The NeRF head predicts densities and semantic logits at each sampled ray location which are used for rendering pixel semantics. Concurrently, we query the Segment-Anything (SAM) foundation model with the camera image to generate a set of unlabeled generic masks. We fuse the masks with the rendered pixel semantics from LiDAR to produce pseudo-labels that supervise the pixel predictions. During inference, we drop the NeRF head and run our model with only LiDAR. We show the effectiveness of our approach in three public LiDAR Semantic Segmentation benchmarks: nuScenes, SemanticKITTI and ScribbleKITTI. Copyright © 2024, The Authors. All rights reserved.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

An Empirical Bayes Analysis of Object Trajectory Representation Models

arXiv

引用

arXiv 2022年

作者： Yao, Yue Goehring, Daniel Reichardt, Joerg Continental AG Germany Freie Universität Berlin Germany Autonomous Vehicles The Dahlem Center for Machine Learning and Robotics The Freie Universität Berlin Germany

Linear trajectory models provide mathematical advantages to autonomous driving applications such as motion prediction. However, linear models’ expressive power and bias for real-world trajectories have not been thoroughly analyzed. We present an in-depth empirical analysis of the trade-off between model complexity and fit error in modelling object trajectories. We analyze vehicle, cyclist, and pedestrian trajectories. Our methodology estimates observation noise and prior distributions over model parameters from several large-scale datasets. Incorporating these priors can then regularize prediction models. Our results show that linear models do represent real-world trajectories with high fidelity at very moderate model complexity. This suggests the feasibility of using linear trajectory models in future motion prediction systems with inherent mathematical advantages. © 2022, CC BY.

关键词： Economic and social effects

来源：评论

学校读者我要写书评

暂无评论

Implementation and Testing of V2X-Applications for Near Future Urban Traffic in Berlin

Implementation and Testing of V2X-Applications for Near Futu...

引用

Omni-layer Intelligent Systems (COINS), International Conference on

作者： Birgit Kwella Kay Massow Bernd Schaeufele Ilja Radusch Johann Nikolai Hark Claas Norman Ritter Bingyi Cao Maximilian Kleinert Christian Wille Oliver Strop Fraunhofer Institute for Open Communication Systems (FOKUS) Berlin Germany Daimler Center for Automotive IT Innovations (DCAITI) TU Berlin Berlin Germany Dahlem Center for Machine Learning and Robotics FU Berlin Berlin Germany Senate Department for Urban Mobility Transport Climate Action and the Environment Berlin Germany Institute of Transportation Systems German Aerospace Center Berlin Germany scope Strop Consulting & Projektmanagement Berlin Germany

ISBN: (数字)9798350349597

ISBN: (纸本)9798350349603

With the emergence of automated vehicles in future urban traffic, the aim is to achieve enhanced connectivity and cooperation among these vehicles themselves, as well as with traffic control systems, to ensure increased traffic safety, even within a mixed operational system. In the project KIS'M, an “AI-based system for connected mobility” is being tested, aimed at addressing the challenges of autonomous, driverless, and on-demand operations in dynamically expanding service areas. In this paper we focus on the importance of interaction between vehicles and cooperative infrastructure for automated and connected driving. In particular, the implementation and testing of cooperative applications in the KIS'M project in Berlin, Germany is presented. The expansion of cooperative infrastructure and related approval processes are addressed, as were various use cases and recommendations for future expansion. The paper emphasizes the need for standardization and technical feasibility to achieve widespread use and improve traffic flow.

关键词： Transportation Standardization Traffic control Safety Intelligent systems Autonomous vehicles Testing

来源：评论

学校读者我要写书评

暂无评论

Robofish as social partner for live guppies 9th

Robofish as social partner for live guppies

引用

9th International Conference on Biomimetic and Biohybrid Systems, Living machines 2020

作者： Musiolek, Lea Hafner, Verena V. Krause, Jens Landgraf, Tim Bierbach, David Department of Computer Science Humboldt-Universität zu Berlin Berlin Germany Faculty of Life Sciences Division of Biology and Ecology of Fishes Humboldt Universität zu Berlin Berlin Germany Dahlem Center for Machine Learning and Robotics Freie Universität Berlin Berlin Germany

ISBN: (纸本)9783030643126

Biomimetic robots that are accepted as social partners by animals may help to gain insights into animals’ social interaction skills. Here, we present an experiment using the biomimetic Robofish which resembles live guppies (Poecilia reticulata) - a small tropical freshwater fish. Guppy females were given the opportunity to interact with different open-loop controlled Robofish replicas. We show that guppies interacting with a lifelike Robofish replica scored higher on social interaction variables than did those faced with a simple white cuboid performing the same movements, although this effect weakened with time. Our study exemplifies the use of Robofish as a research tool, providing highly standardized social cues for the study of fish social skills such as imitation and following. © Springer Nature Switzerland AG 2020.

关键词： Fish

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：