检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Urpí, Núria Armengol Bagatella, Marco Vlastelica, Marin Martius, Georg Department of Computer Science ETH Zurich Zurich Switzerland Max Planck Institute for Intelligent Systems Tübingen Germany Department of Computer Science University of Tübingen Tübingen Germany

Offline data are both valuable and practical resources for teaching robots complex behaviors. Ideally, learning agents should not be constrained by the scarcity of available demonstrations, but rather generalize beyond the training distribution. However, the complexity of real-world scenarios typically requires huge amounts of data to prevent neural network policies from picking up on spurious correlations and learning non-causal relationships. We propose CAIAC, a data augmentation method that can create feasible synthetic transitions from a fixed dataset without having access to online environment interactions. By utilizing principled methods for quantifying causal influence, we are able to perform counterfactual reasoning by swapping action-unaffected parts of the state-space between independent trajectories in the dataset. We empirically show that this leads to a substantial increase in robustness of offline learning algorithms against distributional shift. Videos, code and data are available at https://***/view/caiac. © 2024, CC BY.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

StixelNExT: Toward Monocular Low-Weight Perception for Object Segmentation and Free Space Detection

StixelNExT: Toward Monocular Low-Weight Perception for Objec...

引用

IEEE Symposium on intelligent Vehicle

作者： Marcel Vosshans Omar Ait-Aider Youcef Mezouar Markus Enzweiler Institute for Intelligent Systems Which is Part of the Faculty of Computer Science and Engineering University of Applied Sciences Esslingen Germany Institut Pascal ISPR (Image Systems of Perception Robotics) Universite Clermont Auvergne INP / CNRS France

ISBN: (数字)9798350348811

ISBN: (纸本)9798350348828

In this work, we present a novel approach for general object segmentation from a monocular image, eliminating the need for manually labeled training data and enabling rapid, straightforward training and adaptation with minimal data. Our model initially learns from LiDAR during the training process, which is subsequently removed from the system, allowing it to function solely on monocular imagery. This study leverages the concept of the Stixel-World to recognize a medium level representation of its surroundings. Our network directly predicts a 2D multi-layer Stixel-World and is capable of recognizing and locating multiple, superimposed objects within an image. Due to the scarcity of comparable works, we have divided the capabilities into modules and present a free space detection in our experiments section. Furthermore, we introduce an improved method for generating Stixels from LiDAR data, which we use as ground truth for our network.

关键词： Training Space vehicles Adaptation models Laser radar Image recognition intelligent vehicles Training data

来源：评论

学校读者我要写书评

暂无评论

Dynamic Intent Queries for Motion Transformer-based Trajectory Prediction

arXiv

引用

arXiv 2025年

作者： Demmler, Tobias Hartung, Lennart Tamke, Andreas Dang, Thao Hegai, Alexander Haug, Karsten Mikelsons, Lars Robert Bosch GmbH Stuttgart Germany Institute for Intelligent Systems Department of Computer Science and Engineering Esslingen University Germany University of Augsburg Germany

In autonomous driving, accurately predicting the movements of other traffic participants is crucial, as it significantly influences a vehicle’s planning processes. Modern trajectory prediction models strive to interpret complex patterns and dependencies from agent and map data. The Motion Transformer (MTR) architecture and subsequent work define the most accurate methods in common benchmarks such as the Waymo Open Motion Benchmark. The MTR model employs pre-generated static intention points as initial goal points for trajectory prediction. However, the static nature of these points frequently leads to misalignment with map data in specific traffic scenarios, resulting in unfeasible or unrealistic goal points. Our research addresses this limitation by integrating scene-specific dynamic intention points into the MTR model. This adaptation of the MTR model was trained and evaluated on the Waymo Open Motion Dataset. Our findings demonstrate that incorporating dynamic intention points has a significant positive impact on trajectory prediction accuracy, especially for predictions over long time horizons. Furthermore, we analyze the impact on ground truth trajectories which are not compliant with the map data or are illegal maneuvers. Copyright © 2025, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

PSA: DIFFERENTIALLY PRIVATE STEERING FOR LARGE LANGUAGE MODEL ALIGNMENT

arXiv

引用

arXiv 2025年

作者： Goel, Anmol Hu, Yaxi Gurevych, Iryna Sanyal, Amartya Technical University of Darmstadt Germany Max Planck Institute for Intelligent Systems Tübingen Germany Department of Computer Science University of Copenhagen Denmark

Aligning Large Language Models (LLMs) with human values and away from undesirable behaviors (such as hallucination) has become increasingly important. Recently, steering LLMs towards a desired behavior via activation editing has emerged as an effective method to mitigate harmful generations at inference-time. Activation editing modifies LLM representations by preserving information from positive demonstrations (e.g., truthful) and minimising information from negative demonstrations (e.g., hallucinations). When these demonstrations come from a private dataset, the aligned LLM may leak private information contained in those private samples. In this work, we present the first study of aligning LLM behavior with private datasets. Our work proposes the Private Steering for LLM Alignment (PSA) algorithm to edit LLM activations with differential privacy (DP) guarantees. We conduct extensive experiments on seven different benchmarks with open-source LLMs of different sizes (0.5B to 7B) and model families (LlaMa, Qwen, Mistral and Gemma). Our results show that PSA achieves DP guarantees for LLM alignment with minimal loss in performance, including alignment metrics, open-ended text generation quality, and general-purpose reasoning. We also develop the first Membership Inference Attack (MIA) for evaluating and auditing the empirical privacy for the problem of LLM steering via activation editing. Our experiments support the theoretical guarantees by showing improved guarantees for our PSA algorithm compared to several existing non-private techniques.1 © 2025, CC BY.

关键词： Demonstrations

来源：评论

学校读者我要写书评

暂无评论

Targeted Image Transformation for Improving Robustness in Long Range Aircraft Detection

Targeted Image Transformation for Improving Robustness in Lo...

引用

IEEE/RSJ International Conference on intelligent Robots and systems (IROS)

作者： Rebecca Martin Clement Fung Nikhil Keetha Lujo Bauer Sebastian Scherer School of Computer Science Robotics Institute Carnegie Mellon University Pittsburgh PA USA School of Computer Science Software and Societal Systems Department Carnegie Mellon University Pittsburgh PA USA

ISBN: (数字)9798350377705

ISBN: (纸本)9798350377712

In the field of aviation, the Detect and Avoid (DAA) problem deals with incorporating collision avoidance capabilities into current autopilot navigation systems. As an application of the Small Object Detection (SOD) problem, DAA presents the difficulties of a low signal-to-noise ratio and far range detection. Visual DAA is also susceptible to changing weather and lighting conditions at deployment. While current literature has presented many solutions for this, prior work has yet to study the robustness of the learning-based models for DAA. In this work, we show that standard techniques for improving robustness for object detection do not produce the desired results for DAA given the SOD constraints. We present targeted transformations, a zero-shot technique that can significantly improve robustness with minimal impact on accuracy. We demonstrate how to construct these transformations and evaluate our method on the current SOTA model for DAA, showing a 53.6% increase in recall. This makes our pipeline more robust to changes in lighting and environmental factors, and better able to detect potential threats. In the future, we hope to automate the transformation selection process, making it easier to adopt in different use cases.

关键词： Visualization Navigation Pipelines Lighting LoRa Object detection Robustness Standards Signal to noise ratio Meteorology

来源：评论

学校读者我要写书评

暂无评论

Robust Finite-Time Containment of Networked Heterogeneous Nonlinear systems with Intermittent Measurement Only

引用

IEEE Transactions on Network science and Engineering 2025年

作者： Tian, Biao Zhang, Hao Cui, Peiyu Wang, Zhuping Yan, Huaicheng Shanghai Key Laboratory of Wearable Robotics and Human-Machine Interaction Tongji University College of Electronic and Information Engineering Department of Control Science and Engineering Shanghai201804 China Ministry of Education Shanghai Institute of Intelligent Science and Technology National Key Laboratory of Autonomous Intelligent Unmanned Systems Frontiers Science Center for Intelligent Autonomous Systems China East China University of Science and Technology School of Information Science and Engineering Shanghai200237 China

The robust finite-time containment problem of fully heterogeneous multiagent systems with uncertainties is often challenging, especially when only intermittent output measurement is used. To address the issue posed by the nonidentical system dynamics of multiple leaders, which makes the traditional distributed estimator form for the convex hull inapplicable, a distributed finite-time estimator is constructed for each follower to extract the system matrices and states of leaders. Then, in the output-triggering setting, a finite-time extended state observer driven by intermittent output measurement is developed to reconstruct the uncertainties and unmeasurable states of followers. Meanwhile, the non-continuous output measurement will lead to the differentiation of virtual control laws undefined. To solve this issue, a filtering compensation scheme-based finitetime controller via the backstepping technique is developed for nonlinear followers, ensuring practical finite-time stability (PFTS) of the closed-loop system. It is shown that the proposed algorithm steers each follower into the preset convex combination spanned by the positions of multiple leaders. The capability of the exploited control protocol is verified through simulations and experiments. 2013 IEEE.

关键词： Backstepping

来源：评论

学校读者我要写书评

暂无评论

Clustering of Motion Trajectories by a Distance Measure Based on Semantic Features

Clustering of Motion Trajectories by a Distance Measure Base...

引用

IEEE-RAS International Conference on Humanoid Robots

作者： Christoph Zelch Jan Peters Oskar von Stryk Department of Computer Science TU Darmstadt Hochschulstr. 10 Simulation Systems Optimization and Robotics Group Darmstadt Germany Department of Computer Science TU Darmstadt Hochschulstr. 10 Intelligent Autonomous Systems Group Darmstadt Germany

Clustering of motion trajectories is highly relevant for human-robot interactions as it allows the anticipation of human motions, fast reaction to those, as well as the recognition of explicit gestures. Further, it allows automated analysis of recorded motion data. Many clustering algorithms for trajectories build upon distance metrics that are based on pointwise Euclidean distances. However, our work indicates that focusing on salient characteristics is often sufficient. We present a novel distance measure for motion plans consisting of state and control trajectories that is based on a compressed representation built from their main features. This approach allows a flexible choice of feature classes relevant to the respective task. The distance measure is used in agglomerative hierarchical clustering. We compare our method with the widely used dynamic time warping algorithm on test sets of motion plans for the Furuta pendulum and the Manutec robot arm and on real-world data from a human motion dataset. The proposed method demonstrates slight advantages in clustering and strong advantages in runtime, especially for long trajectories.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Resilient distributed optimization under mobile malicious attacks

引用

IFAC-PapersOnLine 2023年第2期56卷 997-1002页

作者： Yuan Wang Changxin Liu Hideaki Ishii Karl Henrik Johansson Department of Robotics Hunan University Changsha China Division of Decision and Control Systems School of Electrical Engineering and Computer Science KTH Royal Institute of Technology and Digital Futures Stockholm Sweden Department of Computer Science Tokyo Institute of Technology Yokohama Japan

This article addresses the distributed optimization problem in the presence of malicious adversaries that can move within the network and induce faulty behaviors in the attacked nodes. We first investigate the vulnerabilities of a consensus-based secure distributed optimization protocol under mobile adversaries. Then, a modified resilient distributed optimization algorithm is proposed. We develop conditions on the network structure for both complete and non-complete directed graph cases, under which the proposed algorithm guarantees that the estimates by regular nodes converge to the convex combination of the minimizers of their local functions. Simulations are carried out to verify the effectiveness of our approach.

关键词： Fault-tolerant distributed optimization multi-agent systems network security mobile adversary agents

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-Based Prediction of Subsurface Oil Reservoir Pressure Using Spatio-Temporal Data 49

Deep Learning-Based Prediction of Subsurface Oil Reservoir P...

引用

49th Annual Conference of the IEEE Industrial Electronics Society, IECON 2023

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

ISBN: (纸本)9798350331820

Prediction of subsurface oil reservoir pressure are critical to hydrocarbon production. However, the accurate pressure estimation faces great challenges due to the complexity and uncertainty of reservoir. The underground seepage flow and petrophysical parameters (permeability and porosity) are important but difficult to measure in oilfield. Deep learning methods have been successfully used in reservoir engineering and oil & gas production process. In this study, the effective but inaccessible subsurface seepage fields are not used, only the spatial coordinates and temporal information are selected as model input to predict reservoir pressure. A stacked GRU-based deep learning model is proposed to map the relationship between spatio-temporal data and reservoir pressure. The proposed deep learning method is verified by using a three-dimensional reservoir model, and compared with commonly-used methods. The results show that the stacked GRU model has a better performance and higher accuracy than other deep learning or machine learning methods in pressure prediction. © 2023 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Interaction-aware Maneuver Prediction for Autonomous Vehicles using Interaction Graphs

Interaction-aware Maneuver Prediction for Autonomous Vehicle...

引用

IEEE Symposium on intelligent Vehicle

作者： Iago Pachêco Gomes Cristiano Premebida Denis Fernando Wolf Institute of Mathematics and Computer Science University of São Paulo Brazil Department of Electrical and Computer Engineering Institute of Systems and Robotics University of Coimbra Portugal

Intention prediction (IP) is a challenging task for intelligent vehicle’s perception systems. IP provides the likelihood, or probability, of a target vehicle to perform a maneuver subjected to a finite set of possibilities. There are many factors that influence the decision-making process of a driver, which should be considered in a prediction framework. In addition, the lack of labeled large-scale dataset with maneuver annotation imposes another challenge to the task. This paper proposes an Interaction-aware Maneuver Prediction framework, called IAMP, using interaction graphs to extract complex interaction features from traffic scenes. In addition, we explored a semi-supervised approach called Noisy Student to take advantage of unlabeled data in the training step. Experimental results show relevant improvement when using unlabeled data, increasing the average performance of a classifier by 7.17% of accuracy. Moreover, this approach also made it possible to obtain an intention predictor with similar results to a classifier., even when using a shorter observation horizon.

关键词：

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：