检索结果-内蒙古大学图书馆

Fast Spatial Reasoning of Implicit 3D Maps through Explicit Near-Far Sampling Range Prediction

学校读者我要写书评

暂无评论

Fast Spatial Reasoning of Implicit 3D Maps through Explicit ...

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

作者： Chaerin Min Sehyun Cha Changhee Won Jongwoo Lim Department of Computer Science Brown University Advanced Robotics Lab. CTO Division LG Electronics MultiplEYE Co. Ltd Department of Mechanical Engineering Seoul National University

ISBN: (数字)9798350377705

ISBN: (纸本)9798350377712

3D mapping is critical for many robotics applications, such as autonomous navigation and object manipulation. Recently, deep implicit mapping approaches have received much attention for their compactness and ability to represent fine-grained details. However, without explicit guidance, such implicit representations are often cumbersome for searching the full range on the rays to find the object surfaces. As a result, several approaches, including hierarchical sampling, occupancy grids, and zero-level set baking, have been proposed to improve sampling where costly forward passes of the neural network should be performed. However, hierarchical sampling is still suboptimal in that it requires uniform coarse samples. Discrete occupancy grids of Instant NGP and zero-level sets of various baking methods are less suitable for large and noisy real *** this paper, we present a novel framework for adaptively predicting the near-far range for sampling the query positions of the deep implicit map. For this purpose, the truncated signed distance grid for the map is pre-constructed and used to provide hints for near-far prediction during rendering. In addition, our recovery algorithm automatically detects failed near-far predictions and recovers only those rays by directly using the implicit map. We conduct extensive experiments on a synthetic dataset, a public real dataset, and a real dataset captured by our multi-camera robot system. The experimental results show that our algorithm achieves the same rendering quality with surprisingly fewer samples compared to the existing methods, which means that the robot can reason about the image and depth properties of the scene much faster. Finally, a thorough analysis of the sample distribution along the rays is provided to give a better understanding of our method’s strong efficiency, adaptability, and robustness. https://***/TSDF-sampling/

关键词： Three-dimensional displays Neural networks Rendering (computer graphics) Prediction algorithms Search problems Robustness Cognition Noise measurement Intelligent robots Synthetic data

A Pre-study on Data Processing Pipelines for Roadside Object Detection Systems Towards Safer Road Infrastructure

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yu, Yinan Scheidegger, Samuel Grönvall, John-Fredrik Palm, Magnus Svanberg, Erik Wennerby, Johan Amoruso Bakker, Jörg Asymptotic AI SAFER Trafikverket Sweden Volvo Cars Sweden Department of Computer Science and Engineering Chalmers University of Technology Sweden

Single-vehicle accidents are the most common type of fatal accidents in Sweden, where a car drives off the road and runs into hazardous roadside objects. Proper installation and maintenance of protective objects, such as crash cushions and guard rails, may reduce the chance and severity of such accidents. Moreover, efficient detection and management of hazardous roadside objects also plays an important role in improving road safety. To better understand the state-of-the-art and system requirements, in this pre-study, we investigate the feasibility, implementation, limitations and scaling up of data processing pipelines for roadside object detection. In particular, we divide our investigation into three parts: the target of interest, the sensors of choice and the algorithm design. The data sources we consider in this study cover two common setups: 1) road surveying fleet - annual scans conducted by Trafikverket, the Swedish Transport Administration, and 2) consumer vehicle - data collected using a research vehicle from the laboratory of Resource for vehicle research at Chalmers (REVERE). The goal of this report is to investigate how to implement a scalable roadside object detection system towards safe road infrastructure and Sweden’s Vision Zero. © 2022, CC BY-NC-ND.

关键词： Object recognition

Branch-Train-Merge: Embarrassingly Parallel Training of Expert Language Models

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Li, Margaret Gururangan, Suchin Lewis, Mike Althoff, Tim Smith, Noah A. Dettmers, Tim Zettlemoyer, Luke Paul G. Allen School of Computer Science & Engineering University of Washington United States Allen Institute for AI United States Meta AI United States

We present Branch-Train-Merge (BTM), a communication-efficient algorithm for embarrassingly parallel training of large language models (LLMs). We show it is possible to independently train subparts of a new class of LLMs on different subsets of the data, eliminating the massive multi-node synchronization currently required to train LLMs. BTM learns a set of independent EXPERT LMs (ELMs), each specialized to a different textual domain, such as scientific or legal text. These ELMs can be added and removed to update data coverage, ensembled to generalize to new domains, or averaged to collapse back to a single LM for efficient inference. New ELMs are learned by branching from (mixtures of) ELMs in the current set, further training the parameters on data for the new domain, and then merging the resulting model back into the set for future use. Experiments show that BTM improves in- and out-of-domain perplexities as compared to GPT-style Transformer LMs, when controlling for training cost. Through extensive analysis, we show that these results are robust to different ELM initialization schemes, but require expert domain specialization;LM ensembles with random data splits do not perform well. We also present a study of scaling BTM into a new corpus of 64 domains (192B whitespace-separated tokens in total);the resulting LM (22.4B total parameters) performs as well as a Transformer LM trained with 2.5× more compute. These gains grow with the number of domains, suggesting more aggressive parallelism could be used to efficiently train larger models in future work. Copyright © 2022, The Authors. All rights reserved.

关键词： Computational linguistics

SELF-REGULATED BIOLOGICAL TRANSPORTATION STRUCTURES WITH GENERAL ENTROPY DISSIPATION: 2D CASE AND LEAF-SHAPED DOMaiN

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Astuto, Clarissa Markowich, Peter Portaro, Simone Russo, Giovanni Mathematical and Computer Sciences and Engineering Division King Abdullah University of Science and Technology Thuwal23955-6900 Saudi Arabia Department of Mathematics and Computer Sciences University of Catania 95125 Italy Faculty of Mathematics University of Vienna Oskar-Morgenstern-Platz 1 Vienna1090 Austria Department of Mathematics and Computer Sciences University of Catania 95125 Italy

In recent years, the study of biological transportation networks has attracted significant interest, focusing on their self-regulating, demand-driven nature. This paper examines a mathematical model for these networks, featuring nonlinear elliptic equations for pressure and an auxiliary variable, and a reaction-diffusion parabolic equation for the conductivity tensor, introduced in [19]. The model, based on an energy functional with diffusive and metabolic terms, allows for various entropy generating functions, facilitating its application to different biological scenarios. We proved a local well-posedness result for the problem in Hölder spaces employing Schauder and semigroup theory. Then, after a suitable parameter reduction through scaling, we computed the numerical solution for the proposed system using a recently developed ghost nodal finite element method [6]. An interesting aspect emerges when the solution is very articulated and the branches occupy a wide region of the domain. © 2024, CC BY.

关键词： Partial differential equations

A new lightweight data security system for data security in the cloud computing

学校读者我要写书评

暂无评论

Measurement: Sensors 2023年 29卷

作者： Mohammed, Shameer Nanthini, S. Bala Krishna, N. Srinivas, Inumarthi V. Rajagopal, Manikandan Ashok Kumar, M. University of Technology and Applied Sciences Sur Oman Department of AI & DS Saveetha School of Engineering Saveetha Institute of Medical and Technical Science Thandalam Chennai India Department of CSE Mohan Babu University &SreeVidyanikethan Engineering College Andhrapradesh Tirupati517102 India Mumbai400077 India School of Business and Management Christ University Bangalore560029 India Faculty of Computer Science Department of Computer Science and Software Engineering Skyline University Nigeria Kano Nigeria

In recent decades, data has proved indispensable to all facets of human existence. The development of several applications has resulted in the exponential expansion of data. This information can be encrypted and stored in secure areas. Cloud computing is the technology that can be used to store these massive data sets. The article suggests a Cloud-based Data Security System (C-DSS) that employs a five-tiered trust model for cloud-edge data-sharing architectures. The data owner can select an appropriate trust level and Cyber Threat Information (CTI) sanitization procedure before releasing CTI for analytic strategy. In addition, this cleansing method is conducted either by an end device or by the cloud service supplier, based on the organization's degree of confidence in latter. Research presents the trust architecture, cloud architecture, and installation methodology, all of which are intended to meet widest variety of needs for exchanging secret CTI information. The testing findings high degree of data security and an evident improvement in terms of cypher processing time and security services when compared to the encryption systems that are most often employed in cloud technology. In conclusion, research briefly outlines development and evaluation performed so far by pilot applications confirming the architecture. © 2023 The Authors

关键词： Cryptography

IMPROVING EQUIVARIANT NETWORKS WITH PROBABILISTIC SYMMETRY BREAKING

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Lawrence, Hannah Portilheiro, Vasco Zhang, Yan Kaba, Sékou-Oumar Department of Electrical Engineering and Computer Science MIT United States Gatsby Computational Neuroscience Unit UCL United Kingdom Samsung – SAIT AI Lab Montreal Canada Mila – Quebec Artficial Intelligence Institute Canada McGill University Canada

Equivariance encodes known symmetries into neural networks, often enhancing generalization. However, equivariant networks cannot break symmetries: the output of an equivariant network must, by definition, have at least the same self-symmetries as the input. This poses an important problem, both (1) for prediction tasks on domains where self-symmetries are common, and (2) for generative models, which must break symmetries in order to reconstruct from highly symmetric latent spaces. This fundamental limitation can be addressed by considering equivariant conditional distributions, instead of equivariant functions. We present novel theoretical results that establish necessary and sufficient conditions for representing such distributions. Concretely, this representation provides a practical framework for breaking symmetries in any equivariant network via randomized canonicalization. Our method, SymPE (Symmetry-breaking Positional Encodings), admits a simple interpretation in terms of positional encodings. This approach expands the representational power of equivariant networks while retaining the inductive bias of symmetry, which we justify through generalization bounds. Experimental results demonstrate that SymPE significantly improves performance of group-equivariant and graph neural networks across diffusion models for graphs, graph autoencoders, and lattice spin system modeling. © 2025, CC BY-NC-SA.

关键词： Graph neural networks

Real-Time Explosion Dynamics in a Thulium-doped Linear Fiber Laser

学校读者我要写书评

暂无评论

Real-Time Explosion Dynamics in a Thulium-doped Linear Fiber...

CLEO: science and Innovations, S and I 2022

作者： Zeng, Junjie Sander, Michelle Y. Department of Electrical and Computer Engineering BU Photonics Center Boston University BostonMA02215 United States Division of Materials Science and Engineering Boston University BrooklineMA02446 United States Department of Biomedical Engineering Boston University BostonMA02215 United States

Real-time explosion dynamics in a transition chaotic state and a dual-wavelength vector soliton state in a thulium-doped linear fiber laser are analyzed with real-time pulse measurements. © Optica Publishing Grou... 详细信息

ISBN: (纸本)9781557528209

关键词： Thulium

Design and Implementation of an IoT-based Emergency Alert and GPS Tracking System using MQTT and GSM/GPS Module

学校读者我要写书评

暂无评论

Design and Implementation of an IoT-based Emergency Alert an...

Trends in Material science and Inventive Materials (ICTMIM), International Conference on

作者： P. Chinnasamy Pranavdev P. S Chintha Sivakrishnaiah T. Sathiya Irfan Alam Divya Priya Degala Department of Computer Science and Engineering School of Computing Kalasalingam Academy of Research and Education Tamilnadu India Department of MCA Annamacharya University Rajampet India Department of CSE Sona College of Technology Salem Tamilnadu India Cyber Security and Digital Forensics Division SCAI VIT Bhopal University Sehore India Department of Computer Science and Engineering MLR Institute of Technology Hyderabad India

ISBN: (数字)9798331501488

ISBN: (纸本)9798331501495

An optimally optimized emergency response system can prove to be lifesaving by reducing response time during the state of emergency. This paper proposes an IoT-based emergency alert and GPS tracking system implemented on an ESP32 microcontroller and an A9G GSM/GPS module. The system has a continuous log of user position and device battery level and pushes data in intervals through MQTT to the Adafruit IO platform for remote monitoring. To achieve best battery life, the A9G module goes into low power mode in idle mode but keeps GPS monitoring running. One more feature highlighted is the SOS button, which makes an emergency call by sending the user's location as an SMS and calling a pre-configured contact number. The system also maintains data integrity with automatic reconnect of MQTT broker when it loses connection. Low power and battery portable, the system is applicable in personal protection, home health care for elderly, outdoor adventure, and remote area surveillance. The article explains hardware design, software process, MQTT-based data communication, and testing of the performance of the system based on precision, network, and response to emergencies. The results verify reliability and cost saving of the system in real-time monitoring and reporting of emergencies. With its energy-saving and usability design, it helps in the development of IoT-backed safety solutions for real-time tracking and emergency response in case of a need.

关键词： Safety devices Protocols Emergency services Real-time systems Batteries Time factors Older adults Usability Global Positioning System Testing

Uncertainty-Calibrated Test-Time Model Adaptation without Forgetting

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Tan, Mingkui Chen, Guohao Wu, Jiaxiang Zhang, Yifan Chen, Yaofo Zhao, Peilin Niu, Shuaicheng The School of Software Engineering South China University of Technology China Pazhou Laboratory Guangzhou China XVERSE China The School of Computing National University of Singapore Singapore Tencent AI Lab China The School of Computer Science and Engineering Nanyang Technological University Singapore

Test-time adaptation (TTA) seeks to tackle potential distribution shifts between training and testing data by adapting a given model w.r.t. any testing sample. This task is particularly important when the test environment changes frequently. Although some recent attempts have been made to handle this task, we still face two key challenges: 1) prior methods have to perform backpropagation for each test sample, resulting in unbearable optimization costs to many applications;2) while existing TTA solutions can significantly improve the test performance on out-of-distribution data, they often suffer from severe performance degradation on in-distribution data after TTA (known as catastrophic forgetting). To this end, we have proposed an Efficient Anti-Forgetting Test-Time Adaptation (EATA) method which develops an active sample selection criterion to identify reliable and non-redundant samples for test-time entropy minimization. To alleviate forgetting, EATA introduces a Fisher regularizer estimated from test samples to constrain important model parameters from drastic changes. However, in EATA, the adopted entropy loss consistently assigns higher confidence to predictions even when the samples are underlying uncertain, leading to overconfident predictions that underestimate the data uncertainty. To tackle this, we further propose EATA with Calibration (EATA-C) to separately exploit the reducible model uncertainty and the inherent data uncertainty for calibrated TTA. Specifically, we compare the divergence between predictions from the full network and its sub-networks to measure the reducible model uncertainty, on which we propose a test-time uncertainty reduction strategy with divergence minimization loss to encourage consistent predictions instead of overconfident ones. To further re-calibrate predicting confidence on different samples, we utilize the disagreement among predicted labels as an indicator of the data uncertainty. Based on this, we devise a min-max entropy

关键词： Forecasting