Article
Computer Science, Artificial Intelligence
Minh On Vu Ngoc, Edwin Carlinet, Jonathan Fabrizio, Thierry Geraud
Summary: This paper presents a method that combines the Dahu pseudo-distance with edge information in a graph-cut optimization framework, leveraging their complementary strengths. The method achieves better performance in noisy and blurred images compared to other distance-based and graph-cut methods, reducing user effort in object selection.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Information Systems
Yue Song, Hao Tang, Nicu Sebe, Wei Wang
Summary: This research proposes a method to decompose the saliency detection task into two cascaded sub-tasks, detail modeling and body filling, and utilizes novel multi-scale detail attention and body attention blocks for precise feature fusion and performance improvement. Experimental results demonstrate state-of-the-art performances on six public datasets.
ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Yunpeng Ma, Zhihong Yu, Yaqin Zhou, Chang Xu, Dabing Yu
Summary: A visual saliency detection method based on scale invariant feature and stacked denoising autoencoder is proposed. The method utilizes a deep belief network to initialize the parameters of the autoencoder network and uses scale invariant feature to design the loss function for self-training and parameter update. Experimental results demonstrate the method's effectiveness in saliency prediction and object segmentation, outperforming other comparison methods.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Lei Lei, Feng Xi, Shengyao Chen, Zhong Liu
Summary: The study introduces an iterated graph cut (IGC) method for automatic and accurate segmentation of finger-vein images, utilizing structure-specific contextual clues and constraints for improved performance compared to existing approaches. Extensively evaluated on 4 finger-vein databases, the IGC method outperforms state-of-the-art methods, particularly demonstrating significant improvement in average F-measure values across different databases. This research paves the way for fully automatic image segmentation in the field of biometric technologies.
APPLIED INTELLIGENCE
(2021)
Article
Computer Science, Software Engineering
Melissa Kremer, Peter Caruana, Brandon Haworth, Mubbasir Kapadia, Petros Faloutsos
Summary: This study proposes a parametric model and method for generating real-time saliency maps from the perspective of virtual agents. The model aggregates saliency scores from user-defined parameters and outputs a 2D saliency map that can be modulated by an attention field to incorporate 3D information and a character's state of attentiveness.
COMPUTERS & GRAPHICS-UK
(2022)
Article
Computer Science, Artificial Intelligence
Lei Zhu, Xuejing Kang, Lizhu Ye, Anlong Ming
Summary: This paper proposes an ENCUT model that establishes a balanced graph model by adopting a meaningful-loop and a k-step random walk to enhance small object segmentation. The model is further improved by adding a new RWRT that adds local attention to the segmentation of twigs. Experimental results show that the model achieves state-of-the-art performance among NCut-based segmentation models.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2022)
Article
Computer Science, Artificial Intelligence
Priya Mariam Raju, Deepak Mishra, Prerana Mukherjee
Summary: This study introduces an approach that incorporates instance segmentation into object tracking framework, achieving more accurate target tracking through region proposal module and target localization module. Experimental evaluations on multiple benchmark datasets demonstrate a significant improvement in precision, AUC score, and overlap score compared to recent competing trackers.
IMAGE AND VISION COMPUTING
(2021)
Article
Computer Science, Artificial Intelligence
Alan Lukezic, Jiri Matas, Matej Kristan
Summary: In this paper, a discriminative single-shot segmentation tracker called D3S(2) is proposed. It narrows the gap between visual object tracking and video object segmentation by applying two target models with complementary geometric properties. The overall tracking reliability is further increased by decoupling the object and feature scale estimation.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Computer Science, Artificial Intelligence
Abraham Montoya Obeso, Jenny Benois-Pineau, Mireya Sarai Garcia Vazquez, Alejandro Alvaro Ramirez Acosta
Summary: This paper investigates attention mechanisms in Deep Neural Networks (DNNs) and proposes a method that incorporates human visual attention. The proposed approach achieves faster convergence and better performance in image classification tasks compared to global and local automatic attention mechanisms.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Artificial Intelligence
Jun-ying Hu, C. -J. Richard Shi, Jiang-she Zhang
Summary: The YOLO3-SM method improves object detection performance by introducing a saliency map, achieving higher IOU and mAP values compared to YOLO3.
KNOWLEDGE AND INFORMATION SYSTEMS
(2021)
Article
Engineering, Marine
Miaohui Zhang, Baojun Qiao, Ming Xin, Bo Zhang
Summary: This paper presents an automatic ship detection approach in SAR images using phase spectrum, involving sea-land segmentation and ship detection based on phase spectrum. The proposed method is validated to be efficient through experimental results.
JOURNAL OF OCEAN ENGINEERING AND SCIENCE
(2021)
Article
Computer Science, Information Systems
Kaihua Zhang, Yang Wu, Mingliang Dong, Bo Liu, Dong Liu, Qingshan Liu
Summary: This paper proposes an adaptive spatially and high-order semantically modulated deep network framework for object co-segmentation and co-saliency detection. The framework extracts multi-resolution image features and employs adaptive spatial and high-order semantic modulators to highlight co-object regions and learn rich semantic features respectively. Experimental results demonstrate the superior accuracy of the proposed method on multiple benchmark datasets.
IEEE TRANSACTIONS ON MULTIMEDIA
(2023)
Article
Computer Science, Information Systems
Hemant B. Mahajan, Nilesh Uke, Priya Pise, Makarand Shahade, Vandana G. Dixit, Swapna Bhavsar, Sarita D. Deshpande
Summary: This study demonstrates the construction and deployment of a revolutionary framework using computer vision and deep learning to minimize obstacles in real-time Internet of Things (IoT)-enabled robotics applications. By focusing on sensor-captured streams/images and geographical information, the Internet of Robotic Things (IoRT) is enabled to evolve. The framework utilizes efficient computer vision techniques and a deep learning classifier to anticipate and regulate robot motions, providing higher accuracy and reduced prediction duration. The proposed model exhibits improved efficiency and robustness compared to state-of-the-art approaches, with approximately 5% increased overall accuracy and 84% reduced computational complexity.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Xiaoyang Zheng, Xin Tan, Jie Zhou, Lizhuang Ma, Rynson W. H. Lau
Summary: This paper introduces the use of saliency subitizing as weak supervision for salient object detection, with two modules generating and refining saliency masks. Experimental results demonstrate its superiority over other weakly-supervised methods and comparable performance to some fully-supervised methods.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2021)
Article
Engineering, Electrical & Electronic
Li Lingfei, Wu Tieru
Summary: This paper proposes a fully automatic mesh segmentation method that uses the Fiedler vector to reduce computational cost and decrease operation time, finding multiple parts in only one iteration.
CHINESE JOURNAL OF ELECTRONICS
(2021)