Article
Computer Science, Artificial Intelligence
Di Lin, Yi Wang, Lingyu Liang, Ping Li, C. L. Philip Chen
Summary: This paper presents a fine-grained recognition system that incorporates localization, segmentation, alignment, and classification in a unified deep neural network. By proposing a valve linkage function, the system can adaptively compromise errors of classification and alignment, and help update localization and segmentation. Experimental results confirm the effectiveness of the system.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Engineering, Electrical & Electronic
Shijie Wang, Zhihui Wang, Haojie Li, Jianlong Chang, Wanli Ouyang, Qi Tian
Summary: The current limitation of existing fine-grained image recognition methods is the lack of consideration for the complementary relationship and spatial correspondence between low-level details and high-level semantics. To address this issue, this study proposes a new network structure that utilizes accurate semantics to guide the selection of low-level details, making them spatially corresponding and complementary to high-level semantics. Extensive experiments demonstrate that this method achieves the best performance on multiple datasets under the same settings.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Georgios Stratogiannis, Panagiotis Kouris, Georgios Alexandridis, Georgios Siolas, Giorgos Stamou, Andreas Stafylopatis
Summary: This article introduces a novel framework for semantic enrichment of documents by utilizing hierarchical ontological knowledge and classification techniques. The main contributions include the definition of a theoretical model, a method for handling class imbalance, a methodology for assigning semantic labels, and the introduction of novel metrics for evaluating performance. Extensive experiments on popular datasets show promising results and robustness of the approach.
KNOWLEDGE AND INFORMATION SYSTEMS
(2021)
Article
Computer Science, Information Systems
Chunjie Zhang, Da-Han Wang, Haisheng Li
Summary: The paper introduces a discriminative semantic region selection method for fine-grained recognition (DSRS), which selects image regions and predicts their semantic correlations with classes using pre-trained DCNN models. The joint representations are used for classifier training, and experiments demonstrate the superiority of the method.
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION
(2021)
Article
Computer Science, Artificial Intelligence
Xianlin Zhang, Mengling Shen, Xueming Li, Fangxiang Feng
Summary: This paper focuses on the core problems of fine-grained sketch-based image retrieval (FG-SBIR) and proposes a sketch generation model and a novel FG-SBIR model, which have been proven to outperform state-of-the-art baselines in improving retrieval accuracy.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Artificial Intelligence
Ahmad Alobaid, Oscar Corcho
Summary: This paper presents a novel approach to automatically assigning ontology classes to entity columns in tabular data, without the need for external linguistic resources, lookup services, model training, building a model of the knowledge graph beforehand, or human involvement.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Baolong Liu, Qi Zheng, Yabing Wang, Minsong Zhang, Jianfeng Dong, Xun Wang
Summary: In this paper, the authors propose a new model named FeatInter for the challenging task of video-text retrieval. The model enriches video representation by introducing more fine-grained object-level features and considering the visual and semantic features of objects. Visual-semantic interaction and cross-feature interaction are used to enhance object features and frame features. Experimental results on two challenging video datasets demonstrate the effectiveness of the proposed model.
Article
Engineering, Aerospace
Guimin Jia, Junxian LI
Summary: The paper introduces a new strategy for fine-grained semantic verification of radiotelephony read-backs, which is more effective than traditional methods with an average test accuracy of 93.03%.
CHINESE JOURNAL OF AERONAUTICS
(2022)
Article
Engineering, Electrical & Electronic
Luhan Wang, Pengfeng Xiao, Xueliang Zhang, Xinyang Chen
Summary: This study proposes a novel fine-grained adaptation framework that combines global-local alignment and category-level alignment to address the problems in cross-domain segmentation of remote sensing images (RSIs). Experiments demonstrate that local adaptation and category-level adaptation of RSIs are complementary in cross-domain segmentation, and the integrated framework helps achieve outstanding performance for unsupervised domain adaptation (UDA) semantic segmentation of RSIs.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Article
Computer Science, Hardware & Architecture
Federico Piai, Paolo Atzeni, Paolo Merialdo, Divesh Srivastava
Summary: This paper focuses on the task of semantic type discovery over a set of heterogeneous sources. It proposes the iterative RaF-STD solution, which takes advantage of the redundancy of information across sources to cluster and match attributes.
Article
Computer Science, Information Systems
Yanfeng Hu, Xue Qiao, Luo Xing, Chen Peng
Summary: Fine-grained entity typing is a challenging problem in NLP, and existing methods based on attention mechanisms have limitations in effectively extracting discriminative information. To address this, a diversified semantic attention model (DSAM) is proposed to maximize the extraction of distinctive features by integrating multiple levels of attention and constraints, leading to competitive performance in fine-grained entity typing compared to state-of-the-art methods.
Article
Computer Science, Artificial Intelligence
Tiantian Yan, Jian Shi, Haojie Li, Zhongxuan Luo, Zhihui Wang
Summary: Existing methods of fine-grained image recognition focus on learning features from high-resolution inputs, but their performance declines for low quality images. Therefore, a discriminative information restoration and extraction network (DRE-Net) is proposed for low-resolution fine-grained image recognition, and its effectiveness is demonstrated on multiple datasets.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Artificial Intelligence
Xinhai Liu, Zhizhong Han, Yu-Shen Liu, Matthias Zwicker
Summary: The researchers introduced a new fine-grained 3D shape dataset and proposed a novel method named FG3D-Net to address the issue of limited variance among subcategories in the same category. The method captures fine-grained local details of 3D shapes from multiple rendered views, outperforming other state-of-the-art methods in experiments under the fine-grained 3D shape dataset.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2021)
Article
Computer Science, Information Systems
Dongming Zhou, Canlong Zhang, Yanping Tang, Zhixin Li
Summary: This paper proposes a partially aligned network for person re-identification, which uses accurate local features and a local attention network to capture contextual cues, thus improving the accuracy.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Longguang Wang, Yulan Guo, Xiaoyu Dong, Yingqian Wang, Xinyi Ying, Zaiping Lin, Wei An
Summary: This paper studies the sparsity in convolutional neural networks and proposes a generic sparse mask mechanism to improve the inference efficiency. Sparse masks are learned in both data and channel dimensions to dynamically localize and skip redundant computation at a fine-grained level. They develop SMPointSeg, SMSR, and SMStereo based on the sparse mask mechanism for point cloud semantic segmentation, single image super-resolution, and stereo matching tasks, respectively.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Biochemical Research Methods
Fengqiong Huang, James A. Macklin, Hong Cui, Heather A. Cole, Lorena Endara
BMC BIOINFORMATICS
(2015)
Article
Biochemistry & Molecular Biology
Andrew R. Deans, Suzanna E. Lewis, Eva Huala, Salvatore S. Anzaldo, Michael Ashburner, James P. Balhoff, David C. Blackburn, Judith A. Blake, J. Gordon Burleigh, Bruno Chanet, Laurel D. Cooper, Melanie Courtot, Sandor Csoesz, Hong Cui, Wasila Dahdul, Sandip Das, T. Alexander Dececchi, Agnes Dettai, Rui Diogo, Robert E. Druzinsky, Michel Dumontier, Nico M. Franz, Frank Friedrich, George V. Gkouto, Melissa Haendel, Luke J. Harmon, Terry F. Hayamizu, Yongqun He, Heather M. Hines, Nizar Ibrahim, Laura M. Jackson, Pankaj Jaiswal, Christina James-Zorn, Sebastian Koehler, Guillaume Lecointre, Hilmar Lapp, Carolyn J. Lawrence, Nicolas Le Novere, John G. Lundberg, James Macklin, Austin R. Mast, Peter E. Midford, Istvan Miko, Christopher J. Mungall, Anika Oellrich, David Osumi-Sutherland, Helen Parkinson, Martin J. Ramirez, Stefan Richter, Peter N. Robinson, Alan Ruttenberg, Katja S. Schulz, Erik Segerdell, Katja C. Seltmann, Michael J. Sharkey, Aaron D. Smith, Barry Smith, Chelsea D. Specht, R. Burke Squires, Robert W. Thacker, Anne Thessen, Jose Fernandez-Triana, Mauno Vihinen, Peter D. Vize, Lars Vogt, Christine E. Wall, Ramona L. Walls, Monte Westerfeld, Robert A. Wharton, Christian S. Wirkner, James B. Woolley, Matthew J. Yoder, Aaron M. Zorn, Paula M. Mabee
Article
Biochemical Research Methods
Jin Mao, Lisa R. Moore, Carrine E. Blank, Elvis Hsin-Hui Wu, Marcia Ackerman, Sonali Ranade, Hong Cui
BMC BIOINFORMATICS
(2016)
Article
Biochemical Research Methods
Hong Cui, Dongfang Xu, Steven S. Chong, Martin Ramirez, Thomas Rodenhausen, James A. Macklin, Bertram Ludascher, Robert A. Morris, Eduardo M. Soto, Nicolas Mongiardino Koch
BMC BIOINFORMATICS
(2016)
Article
Mathematical & Computational Biology
Carrine E. Blank, Hong Cui, Lisa R. Moore, Ramona L. Walls
JOURNAL OF BIOMEDICAL SEMANTICS
(2016)
Article
Plant Sciences
Lorena Endara, Heather A. Cole, J. Gordon Burleigh, Nathalie S. Nagalingum, James A. Macklin, Jing Liu, Sonali Ranade, Hong Cui
Article
Plant Sciences
Lorena Endara, Hong Cui, J. Gordon Burleigh
APPLICATIONS IN PLANT SCIENCES
(2018)
Article
Computer Science, Information Systems
Jin Mao, Hong Cui
JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY
(2018)
Article
Mathematical & Computational Biology
Limin Zhang, Xingyi Yang, Zuleima Cota, Hong Cui, Bruce Ford, Hsin-liang Chen, James A. Macklin, Anton Reznicek, Julian Starr
Summary: The study evaluates ontology construction methods through two usability studies and finds that Quick Form, Wizard, and WebProtege are preferred by participants, guiding the design of future iterations.
DATABASE-THE JOURNAL OF BIOLOGICAL DATABASES AND CURATION
(2021)
Article
Mathematical & Computational Biology
Hong Cui, Bruce Ford, Julian Starr, Anton Reznicek, Limin Zhang, James A. Macklin
Summary: The survey results indicate a high level of interest among biologists in adopting controlled vocabularies in phenotype publications, particularly in addressing issues of ambiguity and inconsistency in phenotype descriptions. They believe that a new authoring workflow would better reflect the original meaning of data. While controlled vocabularies are widespread, their actual use is not common.
DATABASE-THE JOURNAL OF BIOLOGICAL DATABASES AND CURATION
(2022)
Article
Mathematical & Computational Biology
Hong Cui, Limin Zhang, Bruce Ford, Hsin-liang Cheng, James A. Macklin, Anton Reznicek, Julian Starr
DATABASE-THE JOURNAL OF BIOLOGICAL DATABASES AND CURATION
(2020)
Article
Biodiversity Conservation
Lorena Endara, Anne E. Thessen, Heather A. Cole, Ramona Walls, Georgios Gkoutos, Yujie Cao, Steven S. Chong, Hong Cui
BIODIVERSITY DATA JOURNAL
(2018)
Article
Biodiversity Conservation
Dongfang Xu, Steven S. Chong, Thomas Rodenhausen, Hong Cui
BIODIVERSITY DATA JOURNAL
(2018)
Proceedings Paper
Computer Science, Theory & Methods
Vikas Yadav, Farig Sadeque, Bryan Heidorn, Hong Cui
TRANSFORMING DIGITAL WORLDS, ICONFERENCE 2018
(2018)
Article
Biodiversity Conservation
Hong Cui, James A. Macklin, Joel Sachs, Anton Reznicek, Julian Starr, Bruce Ford, Lyubomir Penev, Hsin-Liang Chen
BIODIVERSITY DATA JOURNAL
(2018)