Article
Computer Science, Artificial Intelligence
Nan Yang, Xin Luan, Huidi Jia, Zhi Han, Xiaofeng Li, Yandong Tang
Summary: This paper addresses three problems in sequential facial image editing: discontinuous editing, inconsistent editing, and irreversible editing. It proposes a novel model to achieve editing continuity, consistency, and reversibility, and provides a sufficient criterion to determine the effectiveness of the model. The experimental results demonstrate that the proposed model enables flexible editing while preserving facial identity.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2023)
Article
Biochemistry & Molecular Biology
Nina Reuven, Julia Adler, Nadav Myers, Yosef Shaul
Summary: This study presents a method to enrich HDR-dependent edited cells in genome editing by using a co-editing strategy. By using repair of a temperature-sensitive mutation as a selectable marker, selection for edited cells is scarless since editing restores the wild-type sequence. This approach offers a rapid, efficient, and scarless method for selecting genome-edited cells requiring HDR.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2021)
Article
Biochemical Research Methods
Yixuan Qiu, Jiebiao Wang, Jing Lei, Kathryn Roeder
Summary: Marker genes, primarily expressed in a single-cell type, are highly correlated in bulk data, allowing for their identification through correlation patterns. A new algorithm has been developed to detect these marker genes by combining published information and bulk transcriptome data, refining the list of known marker genes based on the correlation structure of the bulk data.
Article
Computer Science, Software Engineering
Sean J. Liu, Maneesh Agrawala, Stephen DiVerdi, Aaron Hertzmann
Summary: ZoomShop is a photo editing tool that allows users to adjust the size, position, and foreshortening of scene elements in an image. By selecting depth ranges and deforming the image, users can achieve various compositional effects.
COMPUTER GRAPHICS FORUM
(2022)
Article
Computer Science, Software Engineering
D. Cascaval, M. Shalah, P. Quinn, R. Bodik, M. Agrawala, A. Schulz
Summary: This paper introduces a novel bidirectional editing system for 3D CAD programs. Users can edit the CAD program and manipulate 3D geometry interchangeably, and the system keeps both representations in sync. By implementing an automatically differentiable domain-specific language for CAD programs, our system enables quick optimization on any expressed program, facilitating rapid and interactive exploration of a constrained 3D design space.
COMPUTER GRAPHICS FORUM
(2022)
Article
Quantum Science & Technology
Xi Lu, Hongwei Lin
Summary: The paper presents an improved quantum supersampling algorithm for quantum ray tracing. By replacing the phase estimation method in quantum supersampling with a robust quantum counting scheme, better performance is achieved compared to classical path tracing algorithm and the original form of quantum supersampling, as demonstrated in simulation experiments.
QUANTUM INFORMATION PROCESSING
(2023)
Article
Multidisciplinary Sciences
Weihe Zhong, Ziduo Yang, Calvin Yu-Chian Chen
Summary: The authors develop a graph-to-edits framework called Graph2Edits for retrosynthesis prediction based on graph neural network. This framework combines the two-stage processes of semi-template-based methods, improving the applicability and interpretability of predictions. Evaluated on the USPTO-50k dataset, the model achieves state-of-the-art performance for semi-template-based retrosynthesis with a promising 55.1% top-1 accuracy.
NATURE COMMUNICATIONS
(2023)
Article
Biochemical Research Methods
Lyuba Bozhilova, Javier Pardo-Diaz, Gesine Reinert, Charlotte M. Deane
Summary: Gene co-expression networks can be constructed in various ways, but it's often unclear which method should be preferred. COGENT provides tools to aid in choosing a construction method without the need for external validation data.
Article
Computer Science, Software Engineering
Zhiyuan Zhang, Zhitong Huang, Jing Liao
Summary: This paper proposes the first framework for layout editing of a single image while preserving its visual properties, enabling continuous editing on a single image.
COMPUTER GRAPHICS FORUM
(2023)
Article
Biochemical Research Methods
Zekuan Yu, Rongyao Yan, Yuanyuan Yu, Xiao Ma, Xiao Liu, Jie Liu, Qiushi Ren, Yanye Lu
Summary: A new data augmentation method called Multiple Lesions Insertion (MLI) is proposed to generate diabetic retinopathy (DR) fundus images by inserting real lesions into healthy fundus images. The synthetic images can be generated according to clinical rules and exhibit realistic texture features and rich details. The MLI method outperforms traditional augmentation methods in DR screening tasks.
BIOMEDICAL OPTICS EXPRESS
(2021)
Article
Computer Science, Information Systems
Mona Alghamdi, Asma Cherif, Abdessamad Imine
Summary: Real-time Collaborative Editors (RCEs) allow large groups of collaborators to share and update multimedia documents in real time. EdgeDoc is an edge-based RCE that provides a more effective solution for managing concurrency and collaboration on mobile devices.
PERVASIVE AND MOBILE COMPUTING
(2021)
Article
Computer Science, Artificial Intelligence
Yongqiang Guo
Summary: This paper presents an in-depth study and analysis of the design of art graphics symbolization through a linear regression algorithm. It improves the training efficiency and classification accuracy of the logistic regression algorithm by selecting a suitable range of learning rate and adding modules of data reordering and regularization penalty method. The emotional characteristics and presentation of graphic symbols are explored, along with their impact on human cognitive behavior in the guidance signage system. The significance of emotional expression of graphic symbols in the guidance signage system is summarized.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Geography
Dipto Sarkar, Jennings T. Anderson
Summary: Traditionally, OpenStreetMap has been a VGI project, but in recent years more corporations have enlisted teams of mappers for editing data on OSM. This has led to tensions between corporate editors and non-corporate editors, but the two groups continue to coexist and produce open geospatial data harmoniously, despite the significant growth in the size and contributions of the corporate editing community.
TRANSACTIONS IN GIS
(2022)
Article
Multidisciplinary Sciences
Sebastien Levesque, Diana Mayorga, Jean-Philippe Fiset, Claudia Goupil, Alexis Duringer, Andreanne Loiselle, Eva Bouchard, Daniel Agudelo, Yannick Doyon
Summary: Prime editing allows precise genetic modifications without the need for donor DNA templates, but its efficiency remains a challenge. In this study, a robust co-selection strategy was designed by co-editing the Na+/K+ ATPase, enabling easy genetic modifications and revealing certain patterns in the editing process. This lays the foundation for creating cellular models and developing cell-type specific co-selection strategies.
NATURE COMMUNICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Xiangnan Ren, Jinjiang Li, Zhen Hua, Xinbo Jiang
Summary: The research aims to guide user attention to key objects through co-saliency detection, obtaining the position of ROIs through algorithms and adjusting their saliency to reduce user visual attraction to ROIs effectively.
CAAI TRANSACTIONS ON INTELLIGENCE TECHNOLOGY
(2021)
Article
Computer Science, Information Systems
Jing Bai, Zhiwen Zeng, Tian Wang, Shaobo Zhang, Neal N. Xiong, Anfeng Liu
Summary: In this article, a trust-based active notice task offloading (TANTO) scheme is proposed to provide trust and low-delay task offloading for resource-limited IoT devices in areas with no available communication infrastructure. The main innovations of TANTO include a novel task offloading mechanism, a trust calculation and reasoning method, and an online UAV trajectory optimization algorithm. Experimental results show that TANTO outperforms previous studies in terms of task completion rate, tasks' average completion time, and UAV's flight cost.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Information Systems
Weiwei Fang, Wenyuan Xu, Chongchong Yu, Neal N. Xiong
Summary: The advent of Deep Neural Networks (DNNs) has empowered numerous computer-vision applications, but it is challenging to efficiently deploy and execute DNNs in industrial scenarios due to computational intensity and resource constraints. This article presents EdgeDI, a framework for executing DNN inference in a partitioned, distributed manner on a cluster of Industrial Internet-of-Things (IIoT) devices. EdgeDI achieves high inference performance through model compression based on deep architecture design and distributed inference based on adaptive workload partitioning.
ACM TRANSACTIONS ON INTERNET TECHNOLOGY
(2023)
Article
Computer Science, Theory & Methods
Zichao Xie, Zeyuan Li, Jinsong Gui, Anfeng Liu, Neal N. Xiong, Shaobo Zhang
Summary: The development of 6G has improved the communication capabilities of IoT devices. However, most devices in wireless systems lack the communication capability of 6G and can only communicate through self-organized networks. These devices have low communication, computing, and caching capacities, making them vulnerable to attacks such as traffic-based attacks. To address this, a novel approach using Unmanned Aerial Vehicles (UAVs) and Wavelet Packet Energy Entropy (UWPEE) is proposed to detect traffic-based attacks and determine the trustworthiness of IoT devices. Experimental results show that the UWPEE scheme can effectively identify traffic-based attacks with high accuracy and reduce UAV flight paths.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Computer Science, Information Systems
Hao Li, Jinwei Wang, Neal Xiong, Yi Zhang, Athanasios V. Vasilakos, Xiangyang Luo
Summary: With the rapid rise of urbanization, criminals are able to exploit social networks to spread secret information by using images as a cover. Determining whether these images contain secret information poses a significant challenge to multimedia computing security. Steganalysis methods based on deep learning can effectively identify the presence of secret information in images transmitted on the Internet in urban scenes, thus playing a crucial role in safeguarding national and social security.
ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS
(2023)
Article
Chemistry, Multidisciplinary
Haotian Chen, Sekione Reward Jeremiah, Changhoon Lee, Jong Hyuk Park
Summary: The Industrial Internet of Things (IIoT) combines smart manufacturing and the Internet of Things, improving product quality and reliability through intelligent connection, real-time data processing, collaborative monitoring, and automatic information processing. This paper proposes an automated smart manufacturing framework based on Digital Twin (DT) and Blockchain, where the data used in DT are from cluster generated after blockchain authentication and only accessed and visualized in the cloud when necessary. Simulation results show that the proposed authentication mode is faster than the standard protocol. Additionally, the DT framework for a smart factory deploys the PDQN DRL model, demonstrating higher accuracy, stability, and reliability.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Theory & Methods
Jiayi Yu, Zeyuan Li, Naixue Xiong, Shaobo Zhang, Anfeng Liu, Athanasios V. Vasilakos
Summary: Mobile Crowd Sensing (MCS) platform is a promising computing paradigm that recruits participants to sense and report data for building and providing services to consumers. However, malicious attacks based on false data and uncertain behaviors of participants can harm cybersecurity and reduce the platform's benefit. This paper proposes a Time, Reliability and Truth-aware Online Auction (TRT-OA) mechanism that considers these properties to ensure cybersecurity and maximize platform benefit. The TRT-OA mechanism achieves computational efficiency, budget feasibility, truthfulness, individual rationality, and strategy-proofness, resulting in a 44.06% increase in benefit compared to OMG and TDMC.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Engineering, Electrical & Electronic
Shiwen Zhang, Wang Hu, Wei Liang, Changjian Lei, Neal N. Xiong
Summary: In this study, an efficient intelligent control algorithm based on the back-propagation neural network (BP) is proposed. The STM32F103 is used as the central controller and multiple sensors are utilized to collect environmental information. The algorithm integrates remote control, voice keywords, and buttons for intelligent control. It also includes a motor drive intelligence control algorithm based on BP neural network (MCBP) to improve control accuracy. An application (APP) is developed to display environmental data and enable various intelligent control functions. Extensive experiments confirm the accuracy and efficiency of the proposed algorithm compared to normal control and programmable logic controller control methods.
IET COMMUNICATIONS
(2023)
Article
Computer Science, Information Systems
Xin Liu, Xiaomeng Liu, Neal Xiong, Dan Luo, Gang Xu, Xiubo Chen
Summary: With the development of cloud computing and other modern technologies, collaborative computing between data is increasing, which brings more attention to privacy protection and secure multi-party computation. This paper proposes a secure search protocol based on graph shape, using the cut-choose method and zero-knowledge proof. The proposed protocol achieves secure graph search and matching while resisting malicious attacks. Experimental simulations demonstrate high execution efficiency.
Article
Computer Science, Information Systems
Xin Liu, Jianwei Kong, Dan Luo, Neal Xiong, Gang Xu, Xiubo Chen
Summary: This paper proposes an Intelligent Semi-Honest System (ISHS) for secret matching against malicious adversaries. It designs a secure computation protocol based on the semi-honest model for the secret matching of text strings using a new digital encoding method and an ECC encryption algorithm. It also presents a text string matching protocol under the malicious model that utilizes the cut-and-choose method and zero-knowledge proof to resist malicious behaviors. The correctness and security of the protocol are analyzed, and it is shown to be more efficient and practical compared to existing algorithms. Secure text matching has important engineering applications.
Article
Computer Science, Information Systems
Fengze Cai, Qiang Hu, Renjie Zhou, Neal Xiong
Summary: Relation extraction is a crucial technology for intelligent information extraction, benefiting intelligent communication systems. Existing approaches suffer from the limitation of one-sided sentence embedding in their final prediction vector. This paper proposes an innovative model, REEGAT, which integrates enhanced word embedding from graph neural networks to improve relation extraction performance.
Article
Computer Science, Information Systems
Shiming He, Meng Guo, Zhuozhou Li, Ying Lei, Siyuan Zhou, Kun Xie, Neal N. Xiong
Summary: KPI clustering is important for AIOps when dealing with a large number of KPIs. This approach divides KPIs into classes and applies the same model to detect anomalies or predict outcomes for the KPIs in each class, reducing computational overhead. However, irregular KPIs caused by varying sampling strategies have not been fully addressed. This study proposes an iterative clustering scheme based on matrix factorization to solve the problem of clustering irregular KPIs and achieves higher NMI compared to non-iterative clustering.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Ziqing Xia, Zhangyang Gao, Anfeng Liu, Neal N. Xiong
Summary: In this paper, an asymmetric quorum-based neighbor discovery (AQND) protocol is proposed to reduce delay, improve energy utilization and lifetime, and outperform previous strategies in main performance indicators.
INFORMATION SCIENCES
(2024)
Article
Computer Science, Information Systems
Xiaohuan Liu, Anfeng Liu, Shaobo Zhang, Tian Wang, Neal N. Xiong
Summary: This paper proposes a delay differentiated services routing (DDSR) scheme to reduce the deployment costs for wireless sensor networks (WSNs) with wake-up radio (WuR) functionality, while meeting the delay requirement of forwarding urgent data and maintaining a long lifetime.
INFORMATION SCIENCES
(2024)
Article
Computer Science, Artificial Intelligence
Qinghua Gu, Yan Wang, Peipei Wang, Xuexian Li, Lu Chen, Neal N. Xiong, Di Liu
Summary: This paper proposes a new ensemble clustering method that combines the influence of cluster level and the base clustering level in a unified framework. The method inserts a global weighting strategy into a local ensemble cluster learning framework, improving the robustness and stability of clustering.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)
Article
Engineering, Civil
Run Liu, Anfeng Liu, Zhenzhe Qu, Neal N. Xiong
Summary: This paper proposes a UAV-enabled Computing-Communications Intelligent Offloading (UAV-CCIO) scheme to achieve energy-efficient task offloading in a UAV-enabled edge network. The scheme selects task gathering nodes and utilizes an optimization strategy to reduce the energy consumption of the UAV. Experimental simulations demonstrate that the performance of the UAV-CCIO scheme is superior to existing schemes.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)