Article
Automation & Control Systems
Linhao Li, Zhen Wang, Qinghua Hu, Yongfeng Dong
Summary: The article introduces a new video sparsity model and dictionary learning operation for foreground detection in intelligent video surveillance, showcasing superior performance compared to current techniques in most cases.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Computer Science, Artificial Intelligence
SivaNagiReddy Kalli, T. Suresh, A. Prasanth, T. Muthumanickam, K. Mohanram
Summary: Automatic moving object detection is gaining research interest for its applications in security provision, traffic monitoring, and anomalies detection. A novel Background Modeling mechanism is proposed in this study to improve the accuracy of moving object detection using a Biased Illumination Field Fuzzy C-means algorithm.
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
R. S. Amshavalli, J. Kalaivani
Summary: This paper proposes a fog computing-based smart video surveillance system that provides less latency, network bandwidth, and response time by localizing data to the edges of the network. The proposed system incorporates adaptive key frame extraction and adaptive contour-based background subtraction, to increase the quality of detecting abnormal motions from surveillance video streams.
JOURNAL OF REAL-TIME IMAGE PROCESSING
(2023)
Article
Engineering, Electrical & Electronic
Xue Song, Baohan Xu, Yu-Gang Jiang
Summary: This study proposes an innovative method for video-in-video advertising using multimodal modeling, which focuses on extracting different representations and learning their complementarity to find content similarity between videos and ads. Experimental results demonstrate the effectiveness and user-friendliness of this method.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2021)
Article
Computer Science, Information Systems
Yong Fan, Xiu He, Yiyi Lin, Zhanchuan Cai
Summary: With the development of industrial informatization, video processing technology has become increasingly important. This paper proposes a novel background modeling scheme for surveillance video based on keyframe and particle shape properties. The results of experiments and comparisons with other algorithms demonstrate the superiority of the proposed model, particularly in dealing with stationary objects.
Article
Automation & Control Systems
Chun-Rong Huang, Wei-Yun Huang, Yi-Sheng Liao, Chien-Cheng Lee, Yu-Wei Yeh
Summary: The article proposes a novel content-adaptive resizing framework (CARF) to boost the computation speed of BM methods in high-resolution surveillance videos, achieving better segmentation results. By simultaneously applying two novel downsampling and upsampling layers, the proposed method shows advantages in both quantitative and qualitative results.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2022)
Article
Engineering, Electrical & Electronic
Lirong Wu, Kejie Huang, Haibin Shen, Lianli Gao
Summary: This research proposes a novel video compression scheme that compresses the foreground and background of surveillance video separately, greatly improving compression ratio through adaptive background updating and interpolation module. Experimental results show that the proposed method requires significantly less bpp than the conventional algorithm H.265 for the same PSNR.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2021)
Article
Mathematics, Applied
Jun-Hao Zhuang, Yi-Si Luo, Xi-Le Zhao, Tai-Xiang Jiang, Yi Chang, Jun Liu
Summary: Outdoor video rain streaks removal is a crucial problem in video processing. Traditional methods based on prior knowledge are insufficient for capturing complex structures of real-world videos, while deep learning methods with large model capacities show promise but require abundant training data. We propose an unsupervised method for video rain streaks removal that solely utilizes rainy videos, combining deep neural networks to capture foreground and background components and incorporating total variation regularization for structured rain streaks.
JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS
(2024)
Article
Engineering, Electrical & Electronic
Manoj Purohit, Manvendra Singh, Saralesh Yadav, Adarsh Kumar Singh, Ajay Kumar, Brajesh Kumar Kaushik
Summary: Image-based sensing is an important source of information for various applications such as surveillance and monitoring. The proposed design utilizes visual and infrared sensors for distributed video analytics, allowing continuous processing of image streams to meet real-time requirements, and providing integrated video analytics and motion detection, enhancing the performance of surveillance systems.
IEEE SENSORS JOURNAL
(2022)
Article
Engineering, Electrical & Electronic
Yu Zhao, Dengyan Luo, Fuchun Wang, Han Gao, Mao Ye, Ce Zhu
Summary: In this paper, an end-to-end Unsupervised foreground-background separation based Video Compression neural Networks (UVCNet) is proposed. Our method combines foreground-background separation with video compression module, utilizing the temporal correlation and relatively static background property to achieve online video compression. Experimental results show that UVCNet outperforms state-of-the-art methods with an average improvement of 2.11 dB in Peak Signal-to-Noise Ratio (PSNR) on surveillance datasets.
IEEE TRANSACTIONS ON BROADCASTING
(2023)
Article
Automation & Control Systems
Jianjun Lei, Tianyi Qin, Bo Peng, Wanqing Li, Zhaoqing Pan, Haifeng Shen, Sam Kwong
Summary: In this article, a novel method is proposed to reduce background induced domain shift for adaptive person Re-ID. The method extracts discriminative foreground and background features using a foreground-background joint clustering module and reduces the interference of background with the extraction of discriminative foreground features using an attention-based feature disentanglement module. Experimental results on three widely used person Re-ID benchmarking datasets have shown that the proposed method achieves promising performance compared with the state-of-the-art methods.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Yoonsik Yang, Haksub Kim, Heeseung Choi, Seungho Chae, Ig-Jae Kim
Summary: The study introduces a video synopsis methodology called SSOcT which effectively condenses surveillance videos by dynamically rearranging tubes. By analyzing scene characteristics and determining an effective spatio-temporal structure, the method summarizes input videos efficiently. Experimental results demonstrate the effectiveness and efficiency of the method in summarizing real-world surveillance videos with diverse scene characteristics.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2021)
Article
Computer Science, Artificial Intelligence
Ali Ahrari, Saber Elsayed, Ruhul Sarker, Daryl Essam, Carlos A. Coello Coello
Summary: This study introduces an adaptive multilevel prediction (AMLP) method for detecting and tracking multiple global optima over time. AMLP employs an adaptive mechanism to determine the near-optimal prediction level at each time step and calculates the strength of diversity introduced after a change. Numerical results demonstrate the superiority of AMLP over other prediction methods in dynamic multimodal optimization.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2021)
Article
Engineering, Electrical & Electronic
Donglai Wei, Yang Liu, Xiaoguang Zhu, Jing Liu, Xinhua Zeng
Summary: This paper proposes a novel fusion method and a Multimodal Supervise-Attention enhanced Fusion (MSAF) framework under weak supervision to address the lack of exploration of multimodal data and the ignore of implicit alignment of multimodal features in video anomaly detection. Extensive experiments on four challenging datasets demonstrate the effectiveness of the framework.
IEEE SIGNAL PROCESSING LETTERS
(2022)
Article
Computer Science, Cybernetics
Badri Narayan Subudhi, Manoj Kumar Panda, T. Veerakumar, Vinit Jakhetiya, S. Esakkirajan
Summary: This study proposes a kernel-induced possibilistic fuzzy associated background subtraction scheme for local change detection from a fixed camera captured sequence. The scheme constructs a robust background model and avoids the influence of noisy pixels and outlier points, which shows good performance on multiple evaluation measures.
IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS
(2023)