Multi-Representation Variational Autoencoder Via Iterative Latent Attention And Implicit Differentiation,
2023
Singapore Management University
Multi-Representation Variational Autoencoder Via Iterative Latent Attention And Implicit Differentiation, Nhu Thuat Tran, Hady Wirawan Lauw
Research Collection School Of Computing and Information Systems
Variational Autoencoder (VAE) offers a non-linear probabilistic modeling of user's preferences. While it has achieved remarkable performance at collaborative filtering, it typically samples a single vector for representing user's preferences, which may be insufficient to capture the user's diverse interests. Existing solutions extend VAE to model multiple interests of users by resorting a variant of self-attentive method, i.e., employing prototypes to group items into clusters, each capturing one topic of user's interests. Despite showing improvements, the current design could be more effective since prototypes are randomly initialized and shared across users, resulting in uninformative and non-personalized clusters.To fill the gap, …
Residual Pattern Learning For Pixel-Wise Out-Of-Distribution Detection In Semantic Segmentation,
2023
Singapore Management University
Residual Pattern Learning For Pixel-Wise Out-Of-Distribution Detection In Semantic Segmentation, Y Liu, Choubo Ding, Yu Tian, Guansong Pang, Vasileios Belagiannis, Ian Reid, Gustavo Carneiro
Research Collection School Of Computing and Information Systems
Semantic segmentation models classify pixels into a set of known ("in-distribution") visual classes. When deployed in an open world, the reliability of these models depends on their ability to not only classify in-distribution pixels but also to detect out-of-distribution (OoD) pixels. Historically, the poor OoD detection performance of these models has motivated the design of methods based on model re-training using synthetic training images that include OoD visual objects. Although successful, these re-trained methods have two issues: 1) their in-distribution segmentation accuracy may drop during re-training, and 2) their OoD detection accuracy does not generalise well to new contexts (e.g., …
Feature Prediction Diffusion Model For Video Anomaly Detection,
2023
Singapore Management University
Feature Prediction Diffusion Model For Video Anomaly Detection, Cheng Yan, Shiyu Zhang, Yang Liu, Guansong Pang, Wenjun Wang
Research Collection School Of Computing and Information Systems
Anomaly detection in the video is an important research area and a challenging task in real applications. Due to the unavailability of large-scale annotated anomaly events, most existing video anomaly detection (VAD) methods focus on learning the distribution of normal samples to detect the substantially deviated samples as anomalies. To well learn the distribution of normal motion and appearance, many auxiliary networks are employed to extract foreground object or action information. These high-level semantic features effectively filter the noise from the background to decrease its influence on detection models. However, the capability of these extra semantic models heavily affects the …
Hallucination Detection: Robustly Discerning Reliable Answers In Large Language Models,
2023
Singapore Management University
Hallucination Detection: Robustly Discerning Reliable Answers In Large Language Models, Yuyuan Chen, Qiang Fu, Yichen Yuan, Zhihao Wen, Ge Fan, Dayiheng Liu, Dongmei Zhang, Zhixu Li, Yanghua Xiao
Research Collection School Of Computing and Information Systems
Large language models (LLMs) have gained widespread adoption in various natural language processing tasks, including question answering and dialogue systems. However, a major drawback of LLMs is the issue of hallucination, where they generate unfaithful or inconsistent content that deviates from the input source, leading to severe consequences. In this paper, we propose a robust discriminator named RelD to effectively detect hallucination in LLMs' generated answers. RelD is trained on the constructed RelQA, a bilingual question-answering dialogue dataset along with answers generated by LLMs and a comprehensive set of metrics. Our experimental results demonstrate that the proposed RelD successfully detects …
Dexbert: Effective, Task-Agnostic And Fine-Grained Representation Learning Of Android Bytecode,
2023
Singapore Management University
Dexbert: Effective, Task-Agnostic And Fine-Grained Representation Learning Of Android Bytecode, Tiezhu Sun, Kevin Allix, Kisub Kim, Xin Zhou, Dongsun Kim, David Lo, Tegawendé F. Bissyande, Jacques Klein
Research Collection School Of Computing and Information Systems
The automation of an increasingly large number of software engineering tasks is becoming possible thanks to Machine Learning (ML). One foundational building block in the application of ML to software artifacts is the representation of these artifacts ( e.g. , source code or executable code) into a form that is suitable for learning. Traditionally, researchers and practitioners have relied on manually selected features, based on expert knowledge, for the task at hand. Such knowledge is sometimes imprecise and generally incomplete. To overcome this limitation, many studies have leveraged representation learning, delegating to ML itself the job of automatically devising suitable …
A Gentle Introduction To Chatgpt,
2023
James Madison University
A Gentle Introduction To Chatgpt, Steven W. Holloway
Libraries
A guest lecture on the state of commercial generative transformer technology, mid-2023, to a general audience at Staunton Public Library.
3d Garment Collision Simulation Based On Human Skeletal Features,
2023
School of Information Science and Technology, Beijing Forestry University, Beijing 100083, China
3d Garment Collision Simulation Based On Human Skeletal Features, Yuanyuan Chen, Yongjian Huai, Xiaoying Nie, Ke Lang
Journal of System Simulation
Abstract: In order to enhance the realism of garment and human body collision in real-time fabric simulation, an automated human body fitting collision method based on the bounding box and mesh method is proposed. According to the human skeletal structure and garment type, the skeletal information involved in collision simulation is effectively optimized, so as to better obtain the feature points of the human body and semantically segment them. According to the characteristics of skinning animation, simple capsule colliders and mesh colliders are generated to fit the geometric shape of the human body, and the dynamic following of colliders is …
Research On Support Effectiveness Evaluation Method Of Equipment Systems Based On Pert And Abms,
2023
AVIC Chengdu Aircraft Design & Research Institute, Chengdu 610091, China
Research On Support Effectiveness Evaluation Method Of Equipment Systems Based On Pert And Abms, Shanzhi Ma, Hongliang Wang, Hua He, Weicheng Lun
Journal of System Simulation
Abstract: The support of an equipment system directly affects its combat effectiveness, and the support effectiveness evaluation of equipment systems has the characteristics of large scope, multiple levels, complete elements, and long process. According to the systematic combat requirements of aircraft equipment, the difficulties in evaluating the support effectiveness of aircraft equipment systems are analyzed. The PERT-based modeling method of airfield support is proposed, and the PERT-based process model of equipment system support activity is established according to the modeling requirements and sequential characteristics of aircraft equipment support tasks. The operational model framework of combinable equipment systems based on ABMS …
Simulation Research On Multi-Antenna Coupled Radiation Of Launch Vehicle In Tower,
2023
PLA 63723 Troops, Xinzhou 036300, China
Simulation Research On Multi-Antenna Coupled Radiation Of Launch Vehicle In Tower, Fen Zhang, Tao Yu, Yong Han, Longwei He
Journal of System Simulation
Abstract: The signal radiation of the launch vehicle wireless system test in the closed tower of the launching site is very complex. In order to further study the antenna radiation characteristics, especially the multi-antenna coupled radiation in the whole vehicle state, a multi antenna model with tower-vehicle body is established in this paper based on UG modeling technology and Altair Hyper Works 2017 electromagnetic compatibility simulation platform. It involves the method of moments-physical optics (MOM-PO) hybrid algorithm and delineates different calculation areas for different scale divisions, so as to solve quickly and accurately electromagnetic parameters of multi-antenna coupled radiation. The …
Research On Flight Route Planning For Specific Multi-Missions,
2023
College of Electronic Information, Xijing University, Xi'an 710123, China
Research On Flight Route Planning For Specific Multi-Missions, Lin Zhong, Ming'an Tong, Sheng Li
Journal of System Simulation
Abstract: In order to complete specific aviation missions, a flight route planning model for specific multi-missions is presented. The grid method is used to build a battlefield environment model. Accordingto the complex and real battlefield environment and operational requirements, five target route planning models including distance, fuel consumption, mission completion, ground-to-air threat, and air-to-air threat are proposed. On the basis of specific mission demands, several requirements for missions are analyzed, and the index of mission completion is presented. According to the problem's characteristics, the two-stage solution algorithm for route planning is proposed. In the first stage, the multi-mission sequence is …
Cloud-Edge Collaborative Service Architecture For Lvc Training System,
2023
College of Systems Engineering, National University of Defense Technology, Changsha 410073, China
Cloud-Edge Collaborative Service Architecture For Lvc Training System, Peng Yong, Miao Zhang, Yue Hu
Journal of System Simulation
Abstract: LVC training, an important means of military training, has received great attention from military and M&S experts. As the virtual and physical elements become more abundant and deeply integrated, LVC training systems become increasingly complex. Aiming at physical-virtual connection, information interaction, simulation computation, run-time control, etc., this paper designs a cloud-edge collaborative service architecture for LVC training systems (CESA-LVC) by reference to cyber-physical systems and cloud-edge computing architectures. CESA-LVC standardizes the structures of LVC training systems from several aspects of intelligent real-time interconnection, joint simulation computation, training auxiliary service, training cognitive decision, and dynamic configuration optimization. It provides a …
Rgb-D Saliency Object Detection Based On Cross-Refinement And Circular Attention,
2023
Yunnan University School of Information Science and Engineering, Kuming 650504, China
Rgb-D Saliency Object Detection Based On Cross-Refinement And Circular Attention, Qingqing Dong, Hao Wu, Wenhua Qian, Fengling Kong
Journal of System Simulation
Abstract: In order to solve the problems that the boundary of the saliency object detection area is vague, and the detection area is incomplete or inaccurate, an RGB-D saliency object detection method based on cross-refinement and circular attention is proposed. A cross-refinement module is designed at the stage of extracting features using encoders, which is used to supplement feature information of each other and improve the feature quality before fusion. It also suppresses the negative impact of poor-quality depth maps and addresses the issue that the edges of the saliency object are blurred. For the features after fusion, the circular …
Data Generation Model-Based Synthetic Sample Imputation Method,
2023
Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ), Shenzhen 518107, China; College of Computer Science and Software Engineering, Shenzhen University, Shenzhen 518060, China)
Data Generation Model-Based Synthetic Sample Imputation Method, Yulin He, Jiaqi Chen, Hepeng Xu, Zhexue Huang, Jianfei Yin
Journal of System Simulation
Abstract: In order to solve the problem of inconsistent probability distribution between synthetic samples by imputation and real samples, a data generation model-based synthetic sample imputation (DGM-SSI) method is proposed. The data generation model of real samples is constructed based on the Gaussian mixture model, and the number of corresponding components of the Gaussian mixture model is determined by the multi-model fusion strategy. The synthetic samples required for model imputation are generated by using the data obtained from the real samples. Specifically, the components of the data generation model and their weights are used to control the generation of synthetic …
Research On Hierarchical Motion Planning Method For Uav Substation Inspection,
2023
Department of Automation, North China Electric Power University, Baoding 071003, China
Research On Hierarchical Motion Planning Method For Uav Substation Inspection, Songming Jiao, Yunfeng Shou, Jianpeng Bai, Zhu Wang
Journal of System Simulation
Abstract: In order to improve the efficiency and quality of unmanned aerial vehicle (UAV) substation inspection, a hierarchical motion planning method for UAV inspection based on front-end path search and back-end trajectory generation is proposed. At the front end, an improved A* algorithm is proposed to increase the planning speed and reduce the path turnings by constraining the direction of node expansion and modifying the heuristic function. At the back end, a minimum-snap trajectory optimization combined with the waypoint filtering method is proposed to generate a smooth trajectory that is beneficial for UAV inspection and tracking. The simulation results show …
Aircraft Assignment Method For Optimal Utilization Of Maintenance Intervals,
2023
College of Electronic Information and Automation, Civil Aviation University of China, Tianjin 300300, China
Aircraft Assignment Method For Optimal Utilization Of Maintenance Intervals, Runxia Guo, Yifu Wang
Journal of System Simulation
Abstract: The aircraft assignment problem is studied from a maintenance assurance perspective. In order to ensure its continuous airworthiness, civil aircraft are required to perform maintenance tasks, i. e., scheduled inspections, at specified intervals. The scheduled inspection interval is usually controlled by the number of flight cycles (FC), flight hours (FH), or flight days (FD), whichever comes first. In order to make balanced use of the inspection interval, an aircraft assignment model for a given fleet size is developed to optimize the maintenance interval utilization, and it is solved by a reinforcement learning algorithm to minimize the variance of the …
Fall Detection Method Of Digital Sequence Based On Fusion Strategy,
2023
School of Science, Dalian Jiaotong University, Dalian 116028, China
Fall Detection Method Of Digital Sequence Based On Fusion Strategy, Riming Sun, Hu Guo, Li Zou, Jiaqi Mao, Shengfa Wang
Journal of System Simulation
Abstract: Falls have become the primary cause of disability due to injury for the elderly. Timely and accurate warning of fall events is an important link to rescue work. In order to improve the accuracy of fall detection, a fall detection method based on a fusion strategy is proposed, which considers both the integrity of high-dimensional digital sequences and the specificity of different dimensions. The input digital sequences obtained from the wrist portable sensor are processed by window segmentation according to the saliency of resultant acceleration, so as to ensure the timing of the data and improve the identifiability of …
Virtual Navigation Path Planning Based On Octree Potential Field For Endonasal Endoscope,
2023
School of Artificial Intelligence, Beijing Normal University, Beijing 100875, China
Virtual Navigation Path Planning Based On Octree Potential Field For Endonasal Endoscope, Wenjing Li, Yanlin Luo, Yuhui Wang, Li Zhu
Journal of System Simulation
Abstract: Virtual navigation can intuitively display the internal structure of human tissue from multiple viewpoints. The navigation path planning algorithm is the key to achieving excellent navigation effects. The traditional centerline extraction algorithm can ensure a wide field of view during navigation, but the time efficiency is not high enough on the complex nasal-skull base volume model. To solve the problem, a rapid navigation path planning algorithm based on the octree potential field is proposed. The space outside the obstacles is modeled by an octree, and the octree potential field is constructed by calculating the potential of all the octree …
A Deep Learning-Based Object Detection Framework For Automatic Asphalt Pavement Patch Detection Using Laser Profiling Images,
2023
Technological University Dublin
A Deep Learning-Based Object Detection Framework For Automatic Asphalt Pavement Patch Detection Using Laser Profiling Images, Ibrahim Hassan Syed, Susan Mckeever Dr., Kieran Feighan, David Power, Dympna O'Sullivan
Conference papers
Road maintenance and the early detection of road defects rely on routine pavement inspections. While advanced 3D laser profiling systems have the capability to automatically identify certain types of distress such as cracks and ruts, more complex pavement damage, including patches, often require manual identification. To address this limitation, this study proposes an automated patch detection system that employs object detection techniques. The results demonstrate the ability of object detection models to accurately identify patches in laser profiling images, indicating that the proposed approach has the capability to significantly enhance automation in visual inspection processes. This has the potential for …
Training Simulation Scenario Generation Based On Particle Swarm Optimization,
2023
College of Intelligence Science, National University of Defense Technology, Changsha 410073, China
Training Simulation Scenario Generation Based On Particle Swarm Optimization, Jianxing Gong, Zimu Wang, Qilong Yang
Journal of System Simulation
Abstract: The training effect in the training simulation scenario is not ideal. Therefore, in order to obtain the training simulation scenario with a better training effect, the training simulation scenario is optimized, and a training simulation scenario generation method based on the PSO algorithm is proposed. A fitness function is constructed based on the improved situation assessment method of the power field model, and the ability weight parameters are determined by combining the improved AHP with computer simulation software; by instantiating particles with the attributes of the combat platform, the particle swarm optimization algorithm is improved to solve the optimization …
Survey On Intelligent Wargaming: Tactical & Campaign Wargame And Strategic Game From Game-Theoretic Perspective,
2023
College of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, China
Survey On Intelligent Wargaming: Tactical & Campaign Wargame And Strategic Game From Game-Theoretic Perspective, Junren Luo, Wanpeng Zhang, Fengtao Xiang, Chaoyuan Jiang, Jing Chen
Journal of System Simulation
Abstract: Wargaming is a pre-practice activity to study national security and competition, military conflict and war, crisis management, and other major strategic issues. An intelligent wargaming system needs the ability of artificial intelligence technology. This paper briefly summarizes the research progress of intelligent game, the evolution of wargaming, intelligent wargaming, and strategic gaming methods. From the perspective of game theory, it analyzes the game problem model for intelligent wargaming and sorts out the application mode of intelligent wargaming and the organization mode of a strategic game. An intelligent wargaming service-oriented architecture based on cloud native is proposed. Pre-training method for …
