检索结果-内蒙古大学图书馆

Enhancing Deep Reinforcement Learning-based Robot Navigation Generalization through Scenario Augmentation

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Wang, Shanze Tan, Mingao Yang, Zhibo Wang, Xianghui Shen, Xiaoyu Huang, Hailong Zhang, Wei The Department of Aeronautical and Aviation Engineering Hong Kong Polytechnic University Hong Kong The Ningbo Key Laboratory of Spatial Intelligence and Digital Derivative Institute of Digital Twin Eastern Institute of Technology Ningbo China The Department of Mechanical Engineering National University of Singapore Singapore

This work focuses on enhancing the generalization performance of deep reinforcement learning-based robot navigation in unseen environments. We present a novel data augmentation approach called scenario augmentation, which enables robots to navigate effectively across diverse settings without altering the training scenario. The method operates by mapping the robot’s observation into an imagined space, generating an imagined action based on this transformed observation, and then remapping this action back to the real action executed in simulation. Through scenario augmentation, we conduct extensive comparative experiments to investigate the underlying causes of suboptimal navigation behaviors in unseen environments. Our analysis indicates that limited training scenarios represent the primary factor behind these undesired behaviors. Experimental results confirm that scenario augmentation substantially enhances the generalization capabilities of deep reinforcement learning-based navigation systems. The improved navigation framework demonstrates exceptional performance by producing near-optimal trajectories with significantly reduced navigation time in real-world applications. Copyright © 2025, The Authors. All rights reserved.

关键词： Deep reinforcement learning

Beyond Visibility Limits: A DRL-Based Navigation Strategy for Unexpected Obstacles

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Tan, Mingao Wang, Shanze Huang, Biao Yang, Zhibo Chen, Rongfei Shen, Xiaoyu Zhang, Wei Ningbo Key Laboratory of Spatial Intelligence and Digital Derivative Institute of Digital Twin Eastern Institute of Technology Ningbo China Department of Aeronautical and Aviation Engineering Hong Kong Polytechnic University Hong Kong Department of Mechanical Engineering National University of Singapore Singapore School of Science Harbin Institute of Technology Shenzhen China

Distance-based reward mechanisms in deep reinforcement learning (DRL) navigation systems suffer from critical safety limitations in dynamic environments, frequently resulting in collisions when visibility is restricted. We propose DRL-NSUO, a novel navigation strategy for unexpected obstacles that leverages the rate of change in LiDAR data as a dynamic environmental perception element. Our approach incorporates a composite reward function with environmental change rate constraints and dynamically adjusted weights through curriculum learning, enabling robots to autonomously balance between path efficiency and safety maximization. We enhance sensitivity to nearby obstacles by implementing short-range feature preprocessing of LiDAR data. Experimental results demonstrate that this method significantly improves both robot and pedestrian safety in complex scenarios compared to traditional DRL-based methods. When evaluated on the BARN navigation dataset, our method achieved superior performance with success rates of 94.0% at 0.5 m/s and 91.0% at 1.0 m/s, outperforming conservative obstacle expansion strategies. These results validate DRL-NSUO’s enhanced practicality and safety for human-robot collaborative environments, including intelligent logistics applications. Copyright © 2025, The Authors. All rights reserved.

关键词： Deep reinforcement learning

Multi-Layer Visual Feature Fusion in Multimodal LLMs: Methods, Analysis, and Best Practices

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Lin, Junyan Chen, Haoran Fan, Yue Fan, Yingqi Jin, Xin Su, Hui Fu, Jinlan Shen, Xiaoyu Ningbo Key Laboratory of Spatial Intelligence and Digital Derivative Institute of Digital Twin EIT China Ocean University of China China Zhejiang Gongshang University China Genmo.ai Meituan Inc Engineering Research Center of Chiplet Design and Manufacturing of Zhejiang Province China NUS Singapore

Multimodal Large Language Models (MLLMs) have made significant advancements in recent years, with visual features playing an increasingly critical role in enhancing model performance. However, the integration of multi-layer visual features in MLLMs remains underexplored, particularly with regard to optimal layer selection and fusion strategies. Existing methods often rely on arbitrary design choices, leading to suboptimal outcomes. In this paper, we systematically investigate two core aspects of multi-layer visual feature fusion: (1) selecting the most effective visual layers and (2) identifying the best fusion approach with the language model. Our experiments reveal that while combining visual features from multiple stages improves generalization, incorporating additional features from the same stage typically leads to diminished performance. Furthermore, we find that direct fusion of multi-layer visual features at the input stage consistently yields superior and more stable performance across various configurations. We make all our code publicly available: https://***/EIT-NLP/Layer_Select_Fuse_for_MLLM. © 2025, CC BY.

关键词： Visual BASIC

Empowering Global Voices: A Data-Efficient, Phoneme-Tone Adaptive Approach to High-Fidelity Speech Synthesis

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Geng, Yizhong Xu, Jizhuo Liang, Zeyu Yang, Jinghan Shi, Xiaoyi Shen, Xiaoyu Logic Intelligence Technology Tsinghua University China Beijing University of Posts and Telecommunications China Peking University China Ningbo Key Laboratory of Spatial Intelligence and Digital Derivative Institute of Digital Twin EIT China

Text-to-speech (TTS) technology has achieved impressive results for widely spoken languages, yet many under-resourced languages remain challenged by limited data and linguistic complexities. In this paper, we present a novel methodology that integrates a data-optimized framework with an advanced acoustic model to build high-quality TTS systems for low-resource scenarios. We demonstrate the effectiveness of our approach using Thai as an illustrative case, where intricate phonetic rules and sparse resources are effectively addressed. Our method enables zero-shot voice cloning and improved performance across diverse client applications, ranging from finance to healthcare, education, and law. Extensive evaluations—both subjective and objective—confirm that our model meets state-of-the-art standards, offering a scalable solution for TTS production in data-limited settings, with significant implications for broader industry adoption and multilingual accessibility. All demos are available in https://***/static/thai/***. Copyright © 2025, The Authors. All rights reserved.

关键词： Semantics

MAER-Nav: Bidirectional Motion Learning Through Mirror-Augmented Experience Replay for Robot Navigation

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Wang, Shanze Tan, Mingao Yang, Zhibo Huang, Biao Shen, Xiaoyu Huang, Hailong Zhang, Wei Department of Aeronautical and Aviation Engineering Hong Kong Polytechnic University Hong Kong Ningbo Key Laboratory of Spatial Intelligence and Digital Derivative Institute of Digital Twin Eastern Institute of Technology Ningbo China Department of Mechanical Engineering National University of Singapore Singapore School of Science Harbin Institute of Technology Shenzhen China

Deep Reinforcement Learning (DRL) based navigation methods have demonstrated promising results for mobile robots, but suffer from limited action flexibility in confined spaces. Conventional DRL approaches predominantly learn forward-motion policies, causing robots to become trapped in complex environments where backward maneuvers are necessary for recovery. This paper presents MAER-Nav (Mirror-Augmented Experience Replay for Robot Navigation), a novel framework that enables bidirectional motion learning without requiring explicit failure-driven hindsight experience replay or reward function modifications. Our approach integrates a mirror-augmented experience replay mechanism with curriculum learning to generate synthetic backward navigation experiences from successful trajectories. Experimental results in both simulation and real-world environments demonstrate that MAER-Nav significantly outperforms state-of-the-art methods while maintaining strong forward navigation capabilities. The framework effectively bridges the gap between the comprehensive action space utilization of traditional planning methods and the environmental adaptability of learning-based approaches, enabling robust navigation in scenarios where conventional DRL methods consistently fail. Copyright © 2025, The Authors. All rights reserved.

关键词： Deep reinforcement learning

The fractal or scaling perspective on progressively generated intra-urban clusters from street junctions

学校读者我要写书评

暂无评论

International Journal of digital Earth 2023年第1期16卷 1944-1961页

作者： Biao He Renzhong Guo Minmin Li Ying Jing Zhigang Zhao Wei Zhu Chen Zhang Chengyue Zhang Ding Ma Research Institute for Smart Cities School of Architecture and Urban PlanningShenzhen University&MNR Technology Innovation Center of Territorial&Spatial Big Data&MNR Key Laboratory of Urban Land Resources Monitoring and Simulation&Guangdong-Hong Kong-Macao Joint Laboratory for Smart CitiesShenzhenPeople’s Republic of China Guangdong Laboratory of Artificial Intelligence and Digital Economy(SZ) ShenzhenPeople’s Republic of China Ningbo Institute of Technology School of EconomicsZhejiang UniversityNingboPeople’s Republic of China School of Resource and Environmental Sciences Wuhan UniversityWuhanPeople’s Republic of China

The underlying complexity of urban space can be manifested by its fractal forms and scaling *** paper examines these characteristics at the intra-urban scale through the lens of clustered street junctions(including road ends)in two Chinese metropolitan areas:Beijing and *** derived the cluster sets with Euclidean distance thresholds starting at 100 meters(m)and ending at 1000 m,and outlined each cluster using a concave-hull method to maintain their original irregular *** each delimited cluster,we examined four urban attributes:gross domestic product,number of street nodes,polygon area,and *** analysis revealed that power law distribution applied to almost every cluster set in terms of the four attributes,but varied from one attribute to another or from city to city,represented primarily byfluctuated power law exponents and ht-index values whose profiles along with the cluster growth can effectively characterize the urban ***,we computed the spectrum of intra-urban scaling exponents with cluster size increments,contributing new insights into the allometric relationships between urban configuration and function.

关键词： spatial clustering urban structure urban function power law intra-urban scaling law

BEYOND CONTENT RELEVANCE: EVALUATING INSTRUCTION FOLLOWING IN RETRIEVAL MODELS

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhou, Jianqun Zheng, Yuanlei Chen, Wei Zheng, Qianqian Su, Hui Zhang, Wei Meng, Rui Shen, Xiaoyu Ningbo Key Laboratory of Spatial Intelligence and Digital Derivative Institute of Digital Twin Eastern Institute of Technology Ningbo China Meituan Inc China Salesforce Research United States School of Software Engineering Huazhong University of Science and Technology China Engineering Research Center of Chiplet Design and Manufacturing of Zhejiang Province China

Instruction-following capabilities in LLMs have progressed significantly, enabling more complex user interactions through detailed prompts. However, retrieval systems have not matched these advances, most of them still relies on traditional lexical and semantic matching techniques that fail to fully capture user intent. Recent efforts have introduced instruction-aware retrieval models, but these primarily focus on intrinsic content relevance, which neglects the importance of customized preferences for broader document-level attributes. This study evaluates the instruction-following capabilities of various retrieval models beyond content relevance, including LLM-based dense retrieval and reranking models. We develop InfoSearch, a novel retrieval evaluation benchmark spanning six document-level attributes: Audience, keyword, Format, Language, Length, and Source, and introduce novel metrics – Strict Instruction Compliance Ratio (SICR) and Weighted Instruction Sensitivity Evaluation (WISE) to accurately assess the models’ responsiveness to instructions. Our findings indicate that although fine-tuning models on instruction-aware retrieval datasets and increasing model size enhance performance, most models still fall short of instruction compliance.1 Copyright © 2024, The Authors. All rights reserved.

关键词： Content based retrieval