检索结果-内蒙古大学图书馆

Optimizing Ego Vehicle Trajectory Prediction: The Graph Enhancement Approach

学校读者我要写书评

暂无评论

Optimizing Ego Vehicle Trajectory Prediction: The Graph Enha...

IS and T International Symposium on Electronic Imaging 2024: Autonomous Vehicles and Machines, AVM 2024

作者： Sharma, Sushil Singh, Aryan Sistu, Ganesh Halton, Mark Eising, Ciarán Data-Driven Computer Engineering Group Dept. of Electronic and Computer Engineering University of Limerick Ireland

Predicting the trajectory of an ego vehicle is a critical component of autonomous driving systems. Current state-of-the-art methods typically rely on Deep Neural Networks (DNNs) and sequential models to process front-view images for future trajectory prediction. However, these approaches often struggle with perspective issues affecting object features in the scene. To address this, we advocate for the use of Bird’s Eye View (BEV) perspectives, which offer unique advantages in capturing spatial relationships and object homogeneity. In our work, we leverage Graph Neural Networks (GNNs) and positional encoding to represent objects in a BEV, achieving competitive performance compared to traditional DNN-based methods. While the BEV-based approach loses some detailed information inherent to front-view images, we balance this by enriching the BEV data by representing it as a graph where relationships between the objects in a scene are captured effectively. © 2024, Society for Imaging Science and Technology.

关键词： Deep neural networks

Image Segmentation: Inducing graph-based learning

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Singh, Aryan Van de Ven, Pepijn Eising, Ciarán Denny, Patrick Data-Driven Computer Engineering Group Dept. of Electronic and Computer Engineering University of Limerick Ireland

This study explores the potential of graph neural networks (GNNs) to enhance semantic segmentation across diverse image modalities. We evaluate the effectiveness of a novel GNN-based U-Net architecture on three distinct datasets: PascalVOC, a standard benchmark for natural image segmentation, WoodScape, a challenging dataset of fisheye images commonly used in autonomous driving, introducing significant geometric distortions;and ISIC2016, a dataset of dermoscopic images for skin lesion segmentation. We compare our proposed UNet-GNN model against established convolutional neural networks (CNNs) based segmentation models, including U-Net and U-Net++, as well as the transformer-based SwinUNet. Unlike these methods, which primarily rely on local convolutional operations or global self-attention, GNNs explicitly model relationships between image regions by constructing and operating on a graph representation of the image features. This approach allows the model to capture long-range dependencies and complex spatial relationships, which we hypothesize will be particularly beneficial for handling geometric distortions present in fisheye imagery and capturing intricate boundaries in medical images. Our analysis demonstrates the versatility of GNNs in addressing diverse segmentation challenges and highlights their potential to improve segmentation accuracy in various applications, including autonomous driving and medical image analysis. Code Available at GitHub. © 2025, CC BY.

关键词： Semantic Segmentation

Deformable Convolution Based Road Scene Semantic Segmentation of Fisheye Images in Autonomous Driving

学校读者我要写书评

暂无评论

arXiv 2024年

This study investigates the effectiveness of modern Deformable Convolutional Neural Networks (DCNNs) for semantic segmentation tasks, particularly in autonomous driving scenarios with fisheye images. These images, providing a wide field of view, pose unique challenges for extracting spatial and geometric information due to dynamic changes in object attributes. Our experiments focus on segmenting the WoodScape fisheye image dataset into ten distinct classes, assessing the Deformable Networks’ ability to capture intricate spatial relationships and improve segmentation accuracy. Additionally, we explore different loss functions to address class imbalance issues and compare the performance of conventional CNN architectures with Deformable Convolution-based CNNs, including Vanilla U-Net and Residual U-Net architectures. The significant improvement in mIoU score resulting from integrating Deformable CNNs demonstrates their effectiveness in handling the geometric distortions present in fisheye imagery, exceeding the performance of traditional CNN architectures. This underscores the significant role of Deformable convolution in enhancing semantic segmentation performance for fisheye imagery. © 2024, CC BY.

关键词： Semantic Segmentation

Optimizing Ego Vehicle Trajectory Prediction: The Graph Enhancement Approach

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Sharma, Sushil Singh, Aryan Sistu, Ganesh Halton, Mark Eising, Ciarán Data-Driven Computer Engineering Group Dept. of Electronic and Computer Engineering University of Limerick Ireland

关键词： Deep neural networks

Fisheye Camera and Ultrasonic Sensor Fusion For Near-Field Obstacle Perception in Bird’s-Eye-View

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Das, Arindam Paul, Sudarshan Scholz, Niko Malviya, Akhilesh Kumar Sistu, Ganesh Bhattacharya, Ujjwal Eising, Ciarán The Department of Electronic and Computer Engineering University of Limerick LimerickV94 T9PX Ireland The Data Driven Computer Engineering Research Group University of Limerick LimerickV94 T9PX Ireland The Department of Driving Software and Systems Valeo India Chennai600130 India Valeo Driving Assistance Research Hummendorfer Str. 74 Kronach96317 Germany Valeo Vision Systems Dunmore Road Co. Galway TuamH54 Y276 Ireland Unit Indian Statistical Institute Kolkata700108 India

Accurate obstacle identification represents a fundamental challenge within the scope of near-field perception for autonomous driving. Conventionally, fisheye cameras are frequently employed for comprehensive surround-view perception, including rear-view obstacle localization. However, the performance of such cameras can significantly deteriorate in low-light conditions, during nighttime, or when subjected to intense sun glare. Conversely, cost-effective sensors like ultrasonic sensors remain largely unaffected under these conditions. Therefore, we present, to our knowledge, the first end-to-end multimodal fusion model tailored for efficient obstacle perception in a bird’s-eye-view (BEV) perspective, utilizing fisheye cameras and ultrasonic sensors. Initially, ResNeXt-50 is employed as a set of unimodal encoders to extract features specific to each modality. Subsequently, the feature space associated with the visible spectrum undergoes transformation into BEV. The fusion of these two modalities is facilitated via concatenation. At the same time, the ultrasonic spectrum-based unimodal feature maps pass through content-aware dilated convolution, applied to mitigate the sensor misalignment between two sensors in the fused feature space. Finally, the fused features are utilized by a two-stage semantic occupancy decoder to generate grid-wise predictions for precise obstacle perception. We conduct a systematic investigation to determine the optimal strategy for multimodal fusion of both sensors. We provide insights into our dataset creation procedures, annotation guidelines, and perform a thorough data analysis to ensure adequate coverage of all scenarios. When applied to our dataset, the experimental results underscore the robustness and effectiveness of our proposed multimodal fusion approach. A short video highlighting an overview of this work along with its more qualitative results can be seen at https://***/JmSLBBL9Ruo. © 2024, CC BY.

关键词： Birds

Design of Blockchain based e-Voting System for Vote Requirements

学校读者我要写书评

暂无评论

Design of Blockchain based e-Voting System for Vote Requirem...

5th International Conference on data Mining, Communications and Information Technology, DMCIT 2021

作者： Choi, Seiwoong Kang, Jihun Chung, Kwang Sik Graduate School Dept. of Information Science Korea National Open University Jongno-gu Dongsung-Dong Seoul Korea Republic of BK 21 Data-Driven Science Future Talent Education Research Group Dept of Computer Science and Engineering Korea University Seoul Korea Republic of Dept. of Computer Science Korea National Open University Jongno-gu Dongsung-Dong Seoul Korea Republic of

Voting refers to the submission of an election or an opinion on a specific matter by expressing an intention on the ballot to a certain place. The existing voting method has the burden of counting time and cost. For this reason, research has been conducted to introduce an e-voting system. However, despite the advantages of e-voting, it is not widely used due to the risk of manipulation of voting results and various requirements. Recently, in order to reduce the risk of data manipulation, research is being conducted to apply the blockchain, a technology that guarantees data integrity, to e-voting. Blockchain guarantees the integrity of data, but has a weakness in secrecy. This paper applies the critical encryption technique to the blockchain and satisfies the requirements for voting such as verifiability, anonymity, fairness, non-reusability, competence, safety, transparency, and non-ticketing. We propose a system design and implementation method. The proposed blockchain-based e-voting system provides voter anonymity by issuing a voter certificate based on a blockchain address. The e-voting election monitoring committee generates a threshold group encryption key, and the proposed blockchain-based e-voting system guarantees confidentiality by a threshold group encryption algorithm during the voting process. The voting result is encrypted through a homomorphic encryption algorithm and stored in the blockchain. Thus, the released voting results ensure safety, confidentiality, transparency, and non-vote ticketing. In addition, the proposed blockchain-based e-voting system guarantees the unity and competence of voting through the blockchain's smart contract. © Published under licence by IOP Publishing Ltd.

关键词： Transparency

Survival Analysis of Kasturi Tobacco Plants in Jember District using the Stratified Cox and Extended Cox Models

学校读者我要写书评

暂无评论

MethodsX 2025年

作者： Angga Iryanto Pratama Dwi Rantini Ratih Ardiati Ningrum Arip Ramadan Fazidah Othman Data Science Technology Study Program Faculty of Advanced Technology and Multidiscipline Universitas Airlangga Surabaya 60115 Indonesia Research Group of Data-Driven Decision Support System Faculty of Advanced Technology and Multidiscipline Universitas Airlangga Surabaya 60115 Indonesia Information System Study Program School of Industrial and System Engineering Telkom University Surabaya Campus Surabaya 60231 Indonesia Department of Computer System and Technology Faculty of Computer Science and Information Technology University of Malaya Malaysia

In 2022, tobacco was the top export commodity, generating $106.3 million in sales. East Java Province contributes significantly to Indonesia's tobacco production, with an annual output of 188.6 thousand tons. Jember Regency is the area in East Java Province which is a Kasturi tobacco cultivation area. The productivity of Kasturi tobacco continues to decrease due to several factors that can affect the survival of the tobacco. This study aims to analyze the survival of Kasturi tobacco by creating stratified Cox and extended Cox models in order to handle factors that cannot fulfill the proportional hazard assumption. Based on the results of the analysis, the stratified Cox model is the best model with AIC and BIC values of 798.108 and 805.5748 respectively, while the extended Cox has AIC and BIC values of 840.2186 and 850.1732 respectively. Variables that can significantly affect the survival of Kasturi tobacco are the variable concentration of ZA fertilizer, pesticide concentration. The addition of ZA fertilizer must be appropriate, because if excessive it can cause poisoning of tobacco plants. Likewise with pesticides, if excessive it will cause damage to the leaves. This policy can increase the productivity of Kasturi tobacco. Then, Jember Regency contributes greater export capacity. • This paper aims to determine the factors that affect the survival of Kasturi tobacco plants in Jember • By using the extended Cox model, the best one is obtained using the Heaviside function with AIC and BIC values of 840.2186 and 850.1732, respectively • The stratified Cox model is better than the extended Cox model, which has AIC and BIC values of 798.108 and 805.5748, respectively

关键词： Productive Agriculture Cox Proportional Hazard Stratified Cox Extended Cox Kasturi Tobacco

A two-sided iterative framework for model reduction of linear systems with quadratic output

学校读者我要写书评

暂无评论

A two-sided iterative framework for model reduction of linea...

IEEE Conference on Decision and Control

作者： Ion Victor Gosea Athanasios C. Antoulas Data-Driven Reduction and Identification Group Max Planck Institute for Dynamics of Complex Technical Systems Magdeburg Germany Department of Electrical and Computer Engineering Rice University Houston USA

ISBN: (数字)9781728113982

ISBN: (纸本)9781728113999

We propose a model reduction procedure for approximating large-scale linear systems with quadratic output by means of lower dimensional systems with the same structure. The framework is based on an iteration that, at each step, computes left and right projections matrices (hence two-sided). This is done by means of solving two linear Sylvester equations.

关键词： Linear systems Mathematical model Computational modeling Observability Symmetric matrices Kernel Numerical models

data-driven (Bilinear) identification and reduction

学校读者我要写书评

暂无评论

PAMM 2021年第S1期20卷

作者： Dimitrios S. Karachalios Ion Victor Gosea Athanasios C. Antoulas Max Planck Institute for Dynamics of Complex Technical Systems Data-Driven System Reduction and Identification (DRI) group Sandtorstraße 1 Magdeburg 39106 Max Planck Institute for Dynamics of Complex Technical Systems Data-Driven System Reduction and Identification (DRI) group Sandtorstraße 1 Magdeburg 39106 Rice University Houston Electrical and Computer Engineering Department 6100 Main St. Houston TX 77005 Baylor College of Medicine 1 Baylor Plaza Houston TX 77030

Identifying dynamical systems from measured data is an important step towards accurate modeling and control. Model order reduction (MOR) constitutes a class of methods that can be used to replace large, complex models of dynamical processes with simpler, smaller models. The reduced-order models (ROMs) can be then used for further tasks such as control, design, and simulation. One typical approach for projection-based model reduction for both linear and nonlinear dynamical systems is by employing interpolation. Projection-based methods require access to the internal dynamics of the system which is not always available. The aim here is to compute ROMs without having access to the internal dynamics, by means of a realization independent method. The proposed methodology will fall into the broad category of data-driven approaches. The method under consideration, which will be referred to as the Loewner framework (LF), was originally introduced by the third author. Based on data, LF identifies state-space models in a direct way. In the original setup, the framework relies on compressing the full data set to extract dominant features and, at the same time, to eliminate the inherent redundancies. In the broader class of nonlinear control systems, the LF has been already extended to certain classes with a special structure such as quadratic or bilinear systems. The LF in combination with Volterra series-(VS) offers a non-intrusive approximation method that is capable of identifying bilinear models from input-output time-domain measurements. This method uses harmonic inputs which establish a natural way for data acquisition. The Fourier transform associates these measurements with the derived generalized frequency response functions GFRFs and the LF makes the connection with system theory. In the linear case, the LF associates the so-called S-parameters with the linear transfer function by interpolating in the frequency domain. The goal of the proposed method is to extend ide

关键词：