Article
Computer Science, Artificial Intelligence
Si-Bao Chen, Zi-Han Lin, Chris H. Q. Ding, Bin Luo
Summary: In the field of intelligent transportation and smart city, truck re-identification (Re-ID) plays a crucial role in managing traffic violations. This study introduces a new truck image dataset called Truck-ID, consisting of 32,353 truck images from 7 monitoring sites. To address the specific challenges of truck Re-ID, the dataset is divided into three sub-datasets for comprehensive evaluation. Additionally, an effective Double Granularity Network (DGN) is proposed to integrate global and local features for robust fine-grained truck Re-ID.
COGNITIVE COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Lingbo Liu, Mengmeng Liu, Guanbin Li, Ziyi Wu, Junfan Lin, Liang Lin
Summary: In this work, a novel road-aware traffic flow magnifier (RATFM) is proposed to accurately infer fine-grained traffic flow by explicitly exploiting the prior knowledge of road networks. Extensive experiments show that the proposed RATFM outperforms state-of-the-art models under various scenarios.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Xin Shu, Lei Zhang, Zizhou Wang, Lituan Wang, Zhang Yi
Summary: The core of fine-grained recognition is to distinguish subcategories within a broad category based on subtle differences in images. Two important factors that are less explored are the similarity between categories and the different definitions of fine-grained categories. This paper proposes a multi-granularity classification framework that uses label hierarchies, decoupled and re-coupled features, and a joint probability-based loss to achieve state-of-the-art performance in fine-grained recognition.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Kun Ouyang, Yuxuan Liang, Ye Liu, Zekun Tong, Sijie Ruan, Yu Zheng, David S. Rosenblum
Summary: This paper presents a method for inferring fine-grained urban flows from coarse-grained data, aiming to bridge the gap between storage efficiency and data utility. The proposed method includes an inference network and a fusion subnet, which can effectively generate fine-grained flow distributions. Additionally, the authors propose a cascading model for progressive inference of larger-scale fine-grained urban flows.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Yichao Yan, Ning Zhuang, Bingbing Ni, Jian Zhang, Minghao Xu, Qiang Zhang, Zhang Zheng, Shuo Cheng, Qi Tian, Yi Xu, Xiaokang Yang, Wenjun Zhang
Summary: In this paper, we propose a novel framework GLMGIR for fine-grained team sports video auto-narrative, which uses multi-granular interaction modeling and attention modules to generate continuous linguistic descriptions. We also collect a new video dataset SVN and develop a new evaluation metric FCE to measure the accuracy of the generated descriptions.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Computer Science, Information Systems
Xiao Yu, Hui Liu, Yan Wu, Caiming Zhang
Summary: Recent advancements in multi-view clustering have attracted significant attention. Many existing methods suffer from high time complexity or difficulty in tuning their parameters. Our proposed method, FITS-MSC, introduces a fine-grained similarity fusion strategy to address the issues that may arise from assigning the same weight to instances in partial views.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Hardware & Architecture
Dan Tang, Xiyin Wang, Xiong Li, Pandi Vijayakumar, Neeraj Kumar
Summary: Low-rate denial of service (LDoS) attacks exploit network protocol vulnerabilities to launch periodic bursts, severely impacting TCP application quality of service. Current coarse-scale detection methods are ineffective. To accurately detect LDoS attacks, an adaptive Kohonen Network based fine-grained detection (AKN-FGD) model is proposed. The AKN-FGD scheme achieves accurate detection with high detection performance and adaptability, outperforming other methods.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Ying Yu, Hong Tang, Jin Qian, Zhiliang Zhu, Zhen Cai, Jingqin Lv
Summary: In this paper, we propose an end-to-end trusted multi-granularity information fusion (TMGIF) model for weakly-supervised fine-grained image recognition. By automatically extracting and evaluating the quality of multi-granularity information, and progressively fusing these information, the model is able to generate reliable and interpretable recognition results.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2023)
Article
Computer Science, Interdisciplinary Applications
Wenyan Hu, Stephan Winter, Kourosh Khoshelham
Summary: In this paper, a method for tailored vehicle selection based on forecast fine-grained sensing coverage is proposed without trajectory data. A model is proposed to forecast fine-grained sensing coverage using coarse-grained information of candidate vehicles and a vehicle selection algorithm is developed to maximize the sensing quality. Results show that the selected vehicles based on this method achieve higher sensing quality than two other baselines. This research provides fundamental guidelines for coverage estimation and vehicle selection in urban vehicular sensing applications.
COMPUTERS ENVIRONMENT AND URBAN SYSTEMS
(2023)
Article
Computer Science, Information Systems
Kang Ling, Haipeng Dai, Yuntang Liu, Alex X. Liu, Wei Wang, Qing Gu
Summary: With the rise of AR/VR technology and miniaturization of mobile devices, gesture recognition is becoming popular in the research area of human-computer interaction. This paper presents UltraGesture, an ultrasonic finger motion perception and recognition system that can run on mobile devices without any hardware modification. Experimental results demonstrate that UltraGesture achieves high accuracy in recognizing various gestures.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2022)
Article
Geochemistry & Geophysics
Cheng Zhang, Guoshuai Zhao, Junjiao Liu, Xingjun Zhang, Xueming Qian
Summary: Short-to-medium term temperature prediction in high resolution is a challenging task that requires expertise in various subjects. Our work proposes a fine-grained conditional convolution network (FCCN) to address the limitations of existing methods in modeling meteorological data. Experiments demonstrate that our FCCN model outperforms all other baseline methods in predicting temperature.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Remote Sensing
Dongsheng Chen, Wei Tu, Rui Cao, Yatao Zhang, Biao He, Chisheng Wang, Tiezhu Shi, Qingquan Li
Summary: This paper proposes a hierarchical recognition framework that integrates remote and social sensing data to recognize fine-grained urban villages. Experimental results demonstrate the effectiveness of this framework and a high-precision map is obtained.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2022)
Article
Geochemistry & Geophysics
Zhenqing Wang, Yi Zhou, Futao Wang, Shixin Wang, Gang Qin, Weijie Zou, Zhuochen Wang, Saimiao Liu, Jinfeng Zhu
Summary: This study creates a large-scale high-quality multispectral dataset for fine-grained extraction of buildings and proposes a superior neural network model. The experimental results demonstrate the effectiveness of this method in building extraction and its significance in industries such as disaster response and smart cities.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Computer Science, Information Systems
Kamran Ali, Alex X. Liu
Summary: In this paper, the authors propose VibroTag, a robust and practical vibration based sensing scheme that can extract fine-grained vibration signatures of different surfaces using smartphones with different hardware. The authors implemented VibroTag on two different Android phones and evaluated its performance in various environments, achieving an average accuracy of 86.55 percent while recognizing 24 different locations/surfaces.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2022)
Article
Environmental Sciences
Xuening Qin, Tien Huu Do, Jelle Hofman, Esther Rodrigo Bonet, Valerio Panzica La Manna, Nikos Deligiannis, Wilfried Philips
Summary: Urban air quality mapping plays a crucial role in urban planning, air pollution control, and personal air pollution exposure assessment. Traditional fixed monitoring stations are limited in providing fine-grained air quality maps due to their sparse deployment and inability to capture short-distance variations influenced by factors such as meteorology, road network, and traffic flow. In this study, a context-aware locally adapted deep forest (CLADF) model is proposed to infer the distribution of NO2 with high resolution using measurements from low-cost mobile sensors and contextual factors, particularly traffic flow. The CLADF model outperforms various benchmark models in terms of accuracy and correlation according to extensive validation experiments using mobile NO2 measurements collected in Antwerp, Belgium.