检索结果-内蒙古大学图书馆

CGD-Net: A Hybrid End-to-end Network with gating decoding for Liver Tumor Segmentation from CT Images

学校读者我要写书评

暂无评论

CGD-Net: A Hybrid End-to-end Network with gating decoding fo...

IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)

作者： Xiaogang Zhu Tao Liu Ziqiu Liu Ouyang Shaobo Xin Wang Shu Hu Feng Ding School of Public Policy and Administration Nanchang University School of SoftWare Nanchang University Nanchang University Affiliated Stomatological Hospital School of Public Health University at Albany Department of Computer and Information Technology Purdue University

ISBN: (数字)9798350374285

ISBN: (纸本)9798350374292

Liver tumor segmentation plays a crucial role in the diagnosis and treatment of hepatic lesions. However, accurate tumor segmentation remains a challenging task due to the fuzzy boundaries of liver tumors and the uncertainty in shape, size, and location. In this paper, we propose a new end-to-end segmentation network called CGD-Net, which incorporates Transformer and frequency-domain features into a convolutional network, and proposes a new decoder structure to automatically learn from Segmentation of liver tumors in CT images. The proposed CGD-Net consists of a Transformer encoder, a frequency domain information fusion module, a gated decoder and three skip connections. Using the powerful feature extraction capability of the Transformer encoder to extract multi-feature ***(Control Gate Decoder)blocks gradually restore the feature information lost in the encoding process by emphasizing the original *** order to fully utilize the information of the original image, three skip connections are used to connect each encoder layer and its corresponding decoder layer, and a FEM (Frequency-domain Enhance module)is built in the third skip connection to fuse frequency domain features. Experiments on the LiTS dataset validate that the proposed CGD-Net can effectively segment liver tumors from CT images in an end-to-end manner, with segmentation accuracy exceeding many existing methods.

关键词： Image segmentation Accuracy Computed tomography Frequency-domain analysis Liver Logic gates Transformers

Shaping of Time-Resolved Biphoton Correlations with a Microresonator-Based Spectral Shaper

学校读者我要写书评

暂无评论

Shaping of Time-Resolved Biphoton Correlations with a Micror...

CLEO: Fundamental Science, CLEO:FS 2024 - Part of Conference on Lasers and Electro-Optics, CLEO 2024

作者： Cohen, Lucas M. Wu, Kaiyi Myilswamy, Karthik V. Lingaraju, Navin B. Lu, Hsuan-Hao Lukens, Joseph M. Weiner, Andrew M. School of Electrical and Computer Engineering Purdue University West LafayetteIN47907 United States The Johns Hopkins University Applied Physics Laboratory LaurelMD20723 United States Quantum Information Science Section Oak Ridge National Laboratory Oak RidgeTN37831 United States Research Technology Office and Quantum Collaborative Arizona State University TempeAZ85287 United States

We report on the manipulation of the time-resolved biphoton correlation function using a sub-GHz resolution silicon nitride microresonator-based spectral shaper capable of programmable amplitude and phase modulation. ... 详细信息

ISBN: (纸本)9781957171395

关键词： Silicon nitride

Multi-Robot Cooperative Socially-Aware Navigation Using Multi-Agent Reinforcement Learning

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Weizheng Mao, Le Wang, Ruiqi Min, Byung-Cheol SMART Laboratory Department of Computer and Information Technology Purdue University West LafayetteIN United States College of Mechanical and Electrical Engineering Beijing University of Chemical Technology Beijing China

In public spaces shared with humans, ensuring multi-robot systems navigate without collisions while respecting social norms is challenging, particularly with limited communication. Although current robot social navigation techniques leverage advances in reinforcement learning and deep learning, they frequently overlook robot dynamics in simulations, leading to a simulation-to-reality gap. In this paper, we bridge this gap by presenting a new multi-robot social navigation environment crafted using Dec-POSMDP and multi-agent reinforcement learning. Furthermore, we introduce SAMARL: a novel benchmark for cooperative multi-robot social navigation. SAMARL employs a unique spatial-temporal transformer combined with multi-agent reinforcement learning. This approach effectively captures the complex interactions between robots and humans, thus promoting cooperative tendencies in multi-robot systems. Our extensive experiments reveal that SAMARL outperforms existing baseline and ablation models in our designed environment. Demo videos for this work can be found at: https://***/view/samarl. Copyright © 2023, The Authors. All rights reserved.

关键词： Multipurpose robots

Initial Task Assignment in Multi-Human Multi-Robot Teams: An Attention-enhanced Hierarchical Reinforcement Learning Approach

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Ruiqi Zhao, Dezhong Gupte, Arjun Min, Byung-Cheol SMART Laboratory Department of Computer and Information Technology Purdue University West LafayetteIN United States College of Mechanical and Electrical Engineering Beijing University of Chemical Technology Beijing China

Multi-human multi-robot teams (MH-MR) obtain tremendous potential in tackling intricate and massive missions by merging distinct strengths and expertise of individual members. The inherent heterogeneity of these teams necessitates advanced initial task assignment (ITA) methods that align tasks with the intrinsic capabilities of team members from the outset. While existing reinforcement learning approaches show encouraging results, they might fall short in addressing the nuances of long-horizon ITA problems, particularly in settings with large-scale MH-MR teams or multifaceted tasks. To bridge this gap, we propose an attention-enhanced hierarchical reinforcement learning approach that decomposes the complex ITA problem into structured sub-problems, facilitating more efficient allocations. To bolster sub-policy learning, we introduce a hierarchical cross-attribute attention (HCA) mechanism, encouraging each sub-policy within the hierarchy to discern and leverage the specific nuances in the state space that are crucial for its respective decision-making phase. Through an extensive environmental surveillance case study, we demonstrate the benefits of our model and the HCA inside. Experimental details are available at https://***/view/ita-aehrl. Copyright © 2023, The Authors. All rights reserved.

关键词： Decision making

Real-Time Learning of Wing Motion Correction in an Unconstrained Flapping-Wing Air Vehicle

学校读者我要写书评

暂无评论

Real-Time Learning of Wing Motion Correction in an Unconstra...

IEEE International Conference on Robotic Computing (IRC)

作者： John C. Gallagher Eric T. Matson Ryan Slater Electrical and Computer Enginnering University of Cincinatti Cincinatti OH USA Computer and Information Technology Purdue University West Lafayette IN

ISBN: (纸本)9781665472616

Small Flapping-Wing Micro-Air Vehicles (FW-MAVs) can experience wing damage and wear while in service. Even small amounts of wing can prevent the vehicle from attaining desired waypoints without significant adaptation to onboard flight control. In previous work, we demonstrated that low-level adaptation of wing motion patterns, rather than high-level adaptation of path control, could restore acceptable performance. We further demonstrated that this low-level adaptation could be accomplished while the vehicle was in normal service and without requiring excessive amounts of flight time. Previous work, however, did not carefully consider the use of these methods when the vehicle was completely unconstrained in three-dimensional space (I.E. no mechanical safety supports) and when all vehicle degrees of freedom had to be simultaneously controlled. Also, previous work presumed that the learning algorithm could adapt wing motion patterns with minimal constraints on shape. The newest generation of FW-MAVs we consider place some significant constraints on legal wing motions which brings into question the efficacy of previous work for current vehicles. In this paper, we will provide compelling evidence that learning during unconstrained flight under the newly imposed wing motion conditions is both practical and feasible. This paper constitutes the first formal report of these results and removes the final barriers that had existed to implementation in a fully-realized physical FW-MAV.

关键词： Space vehicles Shape Law Aerospace electronics Real-time systems Safety Robots

NaviSTAR: Socially Aware Robot Navigation with Hybrid Spatio-Temporal Graph Transformer and Preference Learning

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Weizheng Wang, Ruiqi Mao, Le Min, Byung-Cheol SMART Laboratory Department of Computer and Information Technology Purdue University West LafayetteIN United States College of Mechanical and Electrical Engineering Beijing University of Chemical Technology Beijing China

— Developing robotic technologies for use in human society requires ensuring the safety of robots’ navigation behaviors while adhering to pedestrians’ expectations and social norms. However, understanding complex human-robot interactions (HRI) to infer potential cooperation and response among robots and pedestrians for cooperative collision avoidance is challenging. To address these challenges, we propose a novel socially-aware navigation benchmark called NaviSTAR, which utilizes a hybrid Spatio-Temporal grAph tRansformer to understand interactions in human-rich environments fusing crowd multi-modal dynamic features. We leverage an off-policy reinforcement learning algorithm with preference learning to train a policy and a reward function network with supervisor guidance. Additionally, we design a social score function to evaluate the overall performance of social navigation. To compare, we train and test our algorithm with other state-of-the-art methods in both simulator and real-world scenarios independently. Our results show that NaviSTAR outperforms previous methods with outstanding performance1 Copyright © 2023, The Authors. All rights reserved.

关键词： Navigation

Initial Task Allocation for Multi-Human Multi-Robot Teams with Attention-based Deep Reinforcement Learning

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Ruiqi Zhao, Dezhong Min, Byung-Cheol SMART Laboratory Department of Computer and Information Technology Purdue University West LafayetteIN United States College of Mechanical and Electrical Engineering Beijing University of Chemical Technology Beijing China

Multi-human multi-robot teams have great potential for complex and large-scale tasks through the collaboration of humans and robots with diverse capabilities and expertise. To efficiently operate such highly heterogeneous teams and maximize team performance timely, sophisticated initial task allocation strategies that consider individual differences across team members and tasks are required. While existing works have shown promising results in reallocating tasks based on agent state and performance, the neglect of the inherent heterogeneity of the team hinders their effectiveness in realistic scenarios. In this paper, we present a novel formulation of the initial task allocation problem in multi-human multi-robot teams as a contextual multi-attribute decision-make process and propose an attention-based deep reinforcement learning approach. We introduce a cross-attribute attention module to encode the latent and complex dependencies of multiple attributes in the state representation. We conduct a case study in a massive threat surveillance scenario and demonstrate the strengths of our model. Copyright © 2023, The Authors. All rights reserved.

关键词： Reinforcement learning

Self-Supervised Drone Detection Using Acoustic Data

学校读者我要写书评

暂无评论

Self-Supervised Drone Detection Using Acoustic Data

IEEE International Conference on Robotic Computing (IRC)

作者： Juann Kim Mia Y. Wang Eric T. Matson Dept. of Artificial Intelligence Sungkyunkwan University Suwon South Korea Computer Science College of Charleston Charleston USA Computer and Information Technology Purdue University West Lafayette USA

In recent years, the research focus on drone detection and identification via acoustic signals has grown significantly, spanning applications in both commercial and military domains. The predominant approach in this research area involves utilizing labeled data in the supervised machine learning models. However, this approach has inherent limitations as it trains models to detect or identify specific drone types present in the training dataset, constraining their generalization capacity to new drone types and tasks. This paper introduces a novel solution by employing selfsupervised learning to address these limitations. The proposed model, trained without the need for labeled data, exhibits the capability to classify and identify diverse drone types and objects. It learns to differentiate between different drone models by training with a specific drone model, creating representations suitable for various downstream tasks such as classification and identification. Experimental results demonstrate the superiority of this selfsupervised method over state-of-the-art supervised approaches, particularly in the context of drone detection benchmarks that lack label information.

关键词：

QRNN-Transformer: Recognizing Textual Entailment

学校读者我要写书评

暂无评论

QRNN-Transformer: Recognizing Textual Entailment

IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS)

作者： Xiaogang Zhu Zhihan Yan Wenzhan Wang Shu Hu Xin Wang Chunnian Liu School of Public Policy and Administration Nanchang University School of Software Nanchang University Jingdezhen University Department of Computer and Information Technology Purdue University School of Public Health University at Albany State University of New York

ISBN: (数字)9798350374285

ISBN: (纸本)9798350374292

In recent years, the Transformer model based on the self-attention mechanism has made significant progress in natural language processing and has also been applied in the text implication recognition task, achieving excellent results. However, the Transformer model still has deficiencies in modeling local information in the text. To improve the Transformer model, the QRNN-Transformer was proposed, which uses the QRNN network to divide the input text sequence into local short sequences to capture the local information of the input text. The self-attention is improved by combining the gating mechanism to make the model select tasks-related words or features. Extensive experiments have demonstrated the QRNN-Transformer model can effectively improve the accuracy of entailment relationship recognition on both English and Chinese datasets. l

关键词： Accuracy Text recognition Surveillance Transformers Natural language processing Encoding