Article
Optics
Shuanhu Di, Miao Liao, Yuqian Zhao, Yang Li, Yezhan Zeng
Summary: A hierarchical multi-level segmentation framework is proposed in this paper for superpixel segmentation, which starts with initial segmentation using the LI-SLIC method and then hierarchically segments the superpixels to ensure pixels within each superpixel belong to the same object. Finally, the method merges adjacent superpixels based on probability distribution similarity to reduce over-segmentation, showing effectiveness in handling images corrupted by various noises.
OPTICS AND LASER TECHNOLOGY
(2021)
Article
Computer Science, Information Systems
Hua Li, Runmin Cong, Sam Kwong, Chuanbo Chen, Qianqian Xu, Chongyi Li
Summary: An interactive left-right optimization framework for stereo superpixel segmentation is proposed in this paper, achieving superior performance by considering parallax consistency.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Shifeng Li, Yan Cheng, Ye Tian, Yunfeng Liu
Summary: This paper proposes a novel method for detecting abnormal events in videos based on superpixels. By dividing frames into superpixels according to their similarity and compactness, and calculating anomaly scores, the method effectively detects abnormal events in complex scenes.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Xinyu Lin, Yingjie Zhou, Yipeng Liu, Ce Zhu
Summary: This study presents a novel benchmark for assessing CBCD methods, which includes two major contributions. The performance of twelve CBCD methods is evaluated using six distinct metrics based on constructed contour datasets. Experimental results demonstrated that different CBCD algorithms have their positive scenarios and no single method performs the best across all six evaluation metrics.
IMAGE AND VISION COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Sebastian Hoppe Nesgaard Jensen, Mads Emil Brix Doest, Henrik Aanaes, Alessio Del Bue
Summary: Non-rigid structure from motion (nrsfm) is a longstanding and central problem in computer vision, and the lack of high quality data sets has hindered further development in this field. By introducing a large, publicly available data set, this research aims to provide benchmark tools for the advancement of nrsfm techniques.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2021)
Article
Engineering, Electrical & Electronic
Qiuping Jiang, Yuese Gu, Chongyi Li, Runmin Cong, Feng Shao
Summary: This paper discusses the challenges in underwater image enhancement and proposes a new benchmark dataset and a quality assessment metric. The experiments demonstrate the effectiveness of the proposed metric in evaluating the enhanced underwater image quality.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2022)
Article
Business, Finance
K. J. Martijn Cremers, Jon A. Fulkerson, Timothy B. Riley
Summary: This article introduces a new approach to identify benchmark discrepancies in mutual funds and finds that funds with benchmark discrepancies tend to have higher risk than indicated by their prospectus benchmarks.
JOURNAL OF FINANCIAL AND QUANTITATIVE ANALYSIS
(2022)
Article
Microbiology
Kenneth E. E. Schackart III, Jessica B. Graham, Alise J. Ponsero, Bonnie L. Hurwitz
Summary: In this study, 19 metagenomic phage detection tools were evaluated, and 9 were systematically assessed. The results showed significant differences in the predicted phage communities among the different tools, with tools using a homology approach demonstrating lower false positive rates and robustness to eukaryotic contamination, while tools using a sequence composition approach showed higher sensitivity. The benchmark datasets developed in this study are publicly available for future comparability of new tools.
FRONTIERS IN MICROBIOLOGY
(2023)
Article
Chemistry, Multidisciplinary
Weihua Li, Zhuang Miao, Jing Mu, Fanming Li
Summary: Superpixel extraction algorithm based on a seed strategy of contour encoding (SSCE) for infrared images is proposed, which can generate superpixels with high boundary adherence and compactness. The algorithm obtains a contour encoding map by ray scanning the binary edge map and extracts seed points based on a seed sampling strategy. The initial superpixels are generated by clustering and merging similar adjacent superpixels, resulting in improved accuracy and compactness.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Artificial Intelligence
Zaid Al-Huda, Bo Peng, Yan Yang, Riyadh Nazar Ali Algburi
Summary: In this study, a novel approach for improving object segmentation quality is proposed. By selecting optimal segments, using deep seeds, and constructing a graphical model, the quality of object segmentation can be significantly improved without the need for training segmentation networks.
IET IMAGE PROCESSING
(2021)
Article
Computer Science, Artificial Intelligence
Tianshui Chen, Tao Pu, Hefeng Wu, Yuan Xie, Lingbo Liu, Liang Lin
Summary: Facial expression recognition has been a topic of significant research and attention. However, inconsistencies among datasets hinder the generalization ability of learned models. In this study, a unified cross-domain facial expression recognition evaluation benchmark is proposed. A novel adversarial graph representation adaptation framework is developed, which utilizes graph representation propagation and adversarial learning to co-adapt holistic-local features. Extensive evaluations show that the proposed framework outperforms previous state-of-the-art methods.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Computer Science, Hardware & Architecture
Shaofan Wang, Jiaqi Lan, Jing Lin, Yukun Liu, Lichun Wang, Yanfeng Sun, Baocai Yin
Summary: Superpixel segmentation is the process of partitioning an image into sets of pixels that share common visual meanings. This paper proposes an adaptive hypergraph superpixel segmentation method that addresses the issues of ignoring high-order relationships and the difficulty of parameter adjustment. By constructing a hypergraph and encoding the relationships between pixels, the proposed method achieves higher or comparable performance compared to existing methods.
Article
Computer Science, Artificial Intelligence
Jacinto Carrasco, David Lopez, Ignacio Aguilera-Martos, Diego Garcia-Gil, Irina Markova, Marta Garcia-Barzana, Manuel Arias-Rodil, Julian Luengo, Francisco Herrera
Summary: The research in anomaly detection lacks a unified definition of what represents an anomalous instance, leading to diverse paradigms in algorithm design and experimentation. Predictive maintenance represents a special case, where anomalies must be prevented to avoid failures. To address these issues, the concept of positive and negative instances has been generalized into intervals for evaluating unsupervised anomaly detection algorithms.
Article
Environmental Sciences
Lingbo Yang, Limin Wang, Ghali Abdullahi Abubakar, Jingfeng Huang
Summary: High-resolution crop mapping is important for agricultural monitoring and precision agriculture. Combining high-resolution satellite data with medium-resolution time-series images can improve accuracy. The use of the SNIC algorithm for crop mapping based on this combination has shown promising results in terms of classification accuracy, especially when considering the influence of superpixel size.
Article
Computer Science, Interdisciplinary Applications
Jiaxin Yu, Florian Wellmann, Simon Virgo, Marven von Domarus, Mingze Jiang, Joyce Schmatz, Bastian Leibe
Summary: Training data is crucial for the development of ML and DL algorithms. Lack of well-labeled training image data has hindered the development of DL methods like CNNs in mineral thin section image identification. To speed up the annotation process, a human-computer collaborative pipeline using superpixel segmentation is proposed. The pipeline shows efficient and accurate annotation of large and complex images.
COMPUTERS & GEOSCIENCES
(2023)
Article
Engineering, Electrical & Electronic
Xueyu Han, Ishtiaq Rasool Khan, Susanto Rahardja
Summary: This paper proposes a clustering-based TMO method by embedding human visual system models to adapt to different HDR scenes. The method reduces computational complexity using a hierarchical scheme for clustering and enhances local contrast by superimposing details and controlling color saturation by limiting the adaptive saturation parameter. Experimental results show that the proposed method achieves improvements in generating high quality tone-mapped images compared to competing methods.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2024)
Article
Engineering, Electrical & Electronic
Zuopeng Zhao, Tianci Zheng, Kai Hao, Junjie Xu, Shuya Cui, Xiaofeng Liu, Guangming Zhao, Jie Zhou, Chen He
Summary: The research team developed a handheld phone detection network called YOLO-PAI, which successfully achieved real-time detection and underwent testing under various conditions. Experimental results show that YOLO-PAI reduces network structure parameters and computational costs while maintaining accuracy, outperforming other popular networks in terms of speed and accuracy.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2024)
Article
Engineering, Electrical & Electronic
Vivek Sharma, Ashish Kumar Tripathi, Purva Daga, M. Nidhi, Himanshu Mittal
Summary: In this study, a novel ClGan method is proposed for automated plant disease detection. The method reduces the number of parameters and addresses the issues of vanishing gradients, training instability, and non-convergence by using an encoder-decoder network. Additionally, an improved loss function is introduced to stabilize the learning process and optimize weights effectively. Furthermore, a new plant leaf classification method called ClGanNet is introduced, achieving 99.97% training accuracy and 99.04% testing accuracy using the least number of parameters.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2024)
Article
Engineering, Electrical & Electronic
Seongeun Kim, Chang-Ock Lee
Summary: This article introduces a method for segmenting individual teeth in human teeth images by using deep neural networks to obtain pseudo edge-regions and applying active contour models for segmentation.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2024)