Article
Computer Science, Information Systems
Basim Azam, Ranju Mandal, Brijesh Verma
Summary: This research proposes a novel architecture that utilizes distinctive feature selection algorithm and context adaptive information for image parsing tasks, achieving excellent segmentation accuracy on multiple benchmark datasets.
INFORMATION SCIENCES
(2022)
Article
Mathematics
Zhuang Li, Leilei Cao, Hongbin Wang, Lihong Xu
Summary: This paper presents a self-supervised learning method for face parsing, which utilizes unlabeled facial images. By randomly masking patches in the central area of face images, the model is trained to reconstruct the masked patches and capture facial feature representations. The model is then fine-tuned on a small amount of labeled data for improved performance in face parsing.
Article
Computer Science, Artificial Intelligence
Wing-Yin Yu, Lai-Man Po, Yuzhi Zhao, Yujia Zhang, Kin-Wai Lau
Summary: Human parsing is a critical aspect in computer vision, and this paper introduces a novel framework, FEANet, with DenseASPOC context module to address occlusion and small, ambiguous object issues. Experiments show the superiority of FEANet over current methods in various human parsing benchmarks.
IMAGE AND VISION COMPUTING
(2021)
Article
Computer Science, Information Systems
Haiyang Mei, Letian Yu, Ke Xu, Yang Wang, Xin Yang, Xiaopeng Wei, Rynson W. H. Lau
Summary: This article introduces a method for segmenting mirrors and proposes a novel network model called MirrorNet+ to address this problem. The authors construct a large-scale mirror segmentation dataset and conduct extensive experiments to validate the effectiveness and generalization capability of the proposed method. The article also discusses applications of mirror segmentation and possible future research directions.
ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Yutong Gao, Liqian Liang, Congyan Lang, Songhe Feng, Yidong Li, Yunchao Wei
Summary: This work focuses on Interactive Human Parsing (IHP), aiming to segment a human image into multiple body parts based on users' interactions. To address this task, user clicks are used to identify different body parts and transformed into semantic-aware localization maps, which are combined with the RGB image to generate the initial parsing result. The refinement process is explored, and a semantic-perceiving loss is proposed for better optimization. Experimental results show that the IHP solution achieves high parsing accuracy with a small number of user clicks, indicating its potential for data-efficient human parsing.
IEEE TRANSACTIONS ON MULTIMEDIA
(2023)
Article
Computer Science, Artificial Intelligence
Umberto Michieli, Pietro Zanuttigh
Summary: Semantic segmentation of parts of objects is a challenging task. The proposed GMENet addresses this task by combining object-level context conditioning, part-level spatial relationships, and shape contour information. It achieves state-of-the-art results in different benchmarks.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2022)
Article
Environmental Sciences
Xudong Sun, Min Xia, Tianfang Dai
Summary: A remote sensing parsing method combining a controllable fusion module (CFM) and an adaptive edge loss function (AEL) has been proposed, which can significantly improve the parsing effect of satellite images.
Article
Engineering, Multidisciplinary
Hyungjoon Kim, Hyeonwoo Kim, Seongkuk Cho, Eenjun Hwang
Summary: This paper proposes a new face parsing technique that utilizes an attention block combining spatial and channel attention blocks. The structure of these blocks is improved to compensate for their weaknesses. Experimental results demonstrate that the proposed block-based model outperforms other models in segmentation accuracy.
Article
Computer Science, Artificial Intelligence
Zhuo Su, Huiqiang Guan, Yuntian Lai, Fan Zhou, Yun Liang
Summary: This paper proposes a Boundary-guided Part Reasoning Network (BPRNet) based on the Transformer architecture to address the issues of boundary indistinction and parsing inconsistency in human parsing. Experimental results demonstrate that the proposed method outperforms the state-of-the-art methods.
Article
Chemistry, Multidisciplinary
Jun Zhou, Xing Bai, Qin Zhang
Summary: This paper proposes a method to address the issue of misclassification in semantic segmentation models. By calculating the relevance between different class pairs, the method infers the category of connected components and corrects the misclassifications made by deep learning models. Experimental results demonstrate the effectiveness of the method, improving the performance of various deep learning models.
APPLIED SCIENCES-BASEL
(2022)
Article
Chemistry, Multidisciplinary
Jingya Song, Qingxuan Shi, Yihang Li, Fang Yang
Summary: Human parsing is a challenging task in computer vision, involving the accurate segmentation of human body parts. This paper proposes a Global Transformer Module (GTM) and a Detailed Feature Enhancement (DFE) architecture to capture context information and handle small targets, respectively. Experimental results demonstrate the effectiveness of the proposed method.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Civil
Jiaming Zhang, Chaoxiang Ma, Kailun Yang, Alina Roitberg, Kunyu Peng, Rainer Stiefelhagen
Summary: This study introduces panoramic semantic segmentation through the perspective of domain adaptation, establishing a new dataset and framework to address the issue of annotated training data scarcity for panoramic images while achieving unsupervised domain adaptation from conventional pinhole camera images.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Kailun Yang, Xinxin Hu, Rainer Stiefelhagen
Summary: This paper proposes a horizontal and vertical attention module to leverage contextual information in panoramas and presents a multi-source omni-supervised learning scheme for semantic segmentation of wide-FoV images. The Wild PAnoramic Semantic Segmentation (WildPASS) dataset is introduced to facilitate the evaluation of contemporary CNNs in panoramic imagery, reflecting real-world perception challenges in navigation applications.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2021)
Article
Computer Science, Artificial Intelligence
Litao Yu, Zhibin Li, Min Xu, Yongsheng Gao, Jiebo Luo, Jian Zhang
Summary: The Jaccard index, also known as Intersection-over-Union (IoU), is a critical evaluation metric in image semantic segmentation. This paper proposes a margin calibration method as a learning objective to improve the generalization of IoU over data distribution. The method theoretically ensures better segmentation performance in terms of IoU score.
INTERNATIONAL JOURNAL OF COMPUTER VISION
(2022)
Article
Computer Science, Artificial Intelligence
Zhuo Su, Minshi Chen, Enbo Huang, Ge Lin, Fan Zhou
Summary: This study introduces a novel Multi-view Stack Network for human parsing tasks, which effectively utilizes prior information and potential information of image data to address issues such as ambiguous boundaries, incomplete human parts, and redundant labels. Comprehensive experiments on three public datasets demonstrate that the proposed MVSN outperforms existing methods.
Article
Computer Science, Information Systems
Ling Luo, Dingyu Xue, Xinglong Feng
Article
Multidisciplinary Sciences
Xinglong Feng, Xianwen Gao, Ling Luo
Summary: This study proposed a new hot rolled steel strip defect dataset X-SDD for the actual detection of defects on the surface of hot rolled steel strip. Various algorithms were tested on X-SDD, with the results showing that the proposed algorithm achieved high accuracy and outperformed other comparable algorithms.
Article
Multidisciplinary Sciences
Xinglong Feng, Xianwen Gao, Ling Luo