Pub Date : 2024-09-10DOI: 10.1109/TAI.2024.3439048
Hussein Abbass;Keeley Crockett;Jonathan Garibaldi;Alexander Gegov;Uzay Kaymak;Joao Miguel C. Sousa
{"title":"Editorial: From Explainable Artificial Intelligence (xAI) to Understandable Artificial Intelligence (uAI)","authors":"Hussein Abbass;Keeley Crockett;Jonathan Garibaldi;Alexander Gegov;Uzay Kaymak;Joao Miguel C. Sousa","doi":"10.1109/TAI.2024.3439048","DOIUrl":"https://doi.org/10.1109/TAI.2024.3439048","url":null,"abstract":"","PeriodicalId":73305,"journal":{"name":"IEEE transactions on artificial intelligence","volume":"5 9","pages":"4310-4314"},"PeriodicalIF":0.0,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10673750","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142165008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-26DOI: 10.1109/TAI.2024.3450658
Malik Khizar Hayat;Shan Xue;Jia Wu;Jian Yang
Graphs are a foundational way to represent scenarios where objects interact in pairs. Recently, graph neural networks (GNNs) have become widely used for modeling simple graph structures, either in homogeneous or heterogeneous graphs, where edges represent pairwise relationships between nodes. However, many real-world situations involve more complex interactions where multiple nodes interact simultaneously, as observed in contexts such as social groups and gene-gene interactions. Traditional graph embeddings often fail to capture these multifaceted nonpairwise dynamics. A hypergraph, which generalizes a simple graph by connecting two or more nodes via a single hyperedge, offers a more efficient way to represent these interactions. While most existing research focuses on homogeneous and static hypergraph embeddings, many real-world networks are inherently heterogeneous and dynamic. To address this gap, we propose a GNN-based embedding for dynamic heterogeneous hypergraphs, specifically designed to capture nonpairwise interactions and their evolution over time. Unlike traditional embedding methods that rely on distance or meta-path-based strategies for node neighborhood aggregation, a $k$ -hop neighborhood strategy is introduced to effectively encapsulate higher-order interactions in dynamic networks. Furthermore, the information aggregation process is enhanced by incorporating semantic hyperedges, further enriching hypergraph embeddings. Finally, embeddings learned from each timestamp are aggregated using a mean operation to derive the final node embeddings. Extensive experiments on five real-world datasets, along with comparisons against homogeneous, heterogeneous, and hypergraph-based baselines (both static and dynamic), demonstrate the robustness and superiority of our model.
{"title":"Heterogeneous Hypergraph Embedding for Node Classification in Dynamic Networks","authors":"Malik Khizar Hayat;Shan Xue;Jia Wu;Jian Yang","doi":"10.1109/TAI.2024.3450658","DOIUrl":"https://doi.org/10.1109/TAI.2024.3450658","url":null,"abstract":"Graphs are a foundational way to represent scenarios where objects interact in pairs. Recently, graph neural networks (GNNs) have become widely used for modeling simple graph structures, either in homogeneous or heterogeneous graphs, where edges represent pairwise relationships between nodes. However, many real-world situations involve more complex interactions where multiple nodes interact simultaneously, as observed in contexts such as social groups and gene-gene interactions. Traditional graph embeddings often fail to capture these multifaceted nonpairwise dynamics. A hypergraph, which generalizes a simple graph by connecting two or more nodes via a single hyperedge, offers a more efficient way to represent these interactions. While most existing research focuses on homogeneous and static hypergraph embeddings, many real-world networks are inherently heterogeneous and dynamic. To address this gap, we propose a GNN-based embedding for dynamic heterogeneous hypergraphs, specifically designed to capture nonpairwise interactions and their evolution over time. Unlike traditional embedding methods that rely on distance or meta-path-based strategies for node neighborhood aggregation, a \u0000<inline-formula><tex-math>$k$</tex-math></inline-formula>\u0000-hop neighborhood strategy is introduced to effectively encapsulate higher-order interactions in dynamic networks. Furthermore, the information aggregation process is enhanced by incorporating semantic hyperedges, further enriching hypergraph embeddings. Finally, embeddings learned from each timestamp are aggregated using a mean operation to derive the final node embeddings. Extensive experiments on five real-world datasets, along with comparisons against homogeneous, heterogeneous, and hypergraph-based baselines (both static and dynamic), demonstrate the robustness and superiority of our model.","PeriodicalId":73305,"journal":{"name":"IEEE transactions on artificial intelligence","volume":"5 11","pages":"5465-5477"},"PeriodicalIF":0.0,"publicationDate":"2024-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142600204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-14DOI: 10.1109/TAI.2024.3443789
Fok Hing Chi Tivive;Abdesselam Bouzerdoum;Son Lam Phung;Hoang Thanh Le;Hamza Baali
Image-based crowd counting has gained significant attention due to its widespread applications in security and surveillance. Recent advancements in deep learning have led to the development of numerous methods that have achieved remarkable success in accurately counting crowds. However, many of the existing deep learning methods, which have large model sizes, are unsuitable for deployment on edge devices. This article introduces a novel network architecture and processing element designed to create an efficient and compact deep learning model for crowd counting. The processing element, referred to as the shunting inhibitory neuron, generates complex decision boundaries, making it more powerful than the traditional perceptron. It is employed in both the encoder and decoder modules of the proposed model for feature extraction. Furthermore, the decoder includes alternating convolutional and transformer layers, which provide local receptive fields and global self-attention, respectively. This design captures rich contextual information that is used for generating accurate segmentation and density maps. The self-attention mechanism is implemented using convolution modulation instead of matrix multiplication to reduce computational costs. Experiments conducted on three challenging crowd counting datasets demonstrate that the proposed deep learning network, which comprises a small model size, achieves crowd counting performance comparable to that of state-of-the-art techniques. Codes are available at https://github.com/ftivive/SINet .
{"title":"A Deep Learning-Based Method for Crowd Counting Using Shunting Inhibition Mechanism","authors":"Fok Hing Chi Tivive;Abdesselam Bouzerdoum;Son Lam Phung;Hoang Thanh Le;Hamza Baali","doi":"10.1109/TAI.2024.3443789","DOIUrl":"https://doi.org/10.1109/TAI.2024.3443789","url":null,"abstract":"Image-based crowd counting has gained significant attention due to its widespread applications in security and surveillance. Recent advancements in deep learning have led to the development of numerous methods that have achieved remarkable success in accurately counting crowds. However, many of the existing deep learning methods, which have large model sizes, are unsuitable for deployment on edge devices. This article introduces a novel network architecture and processing element designed to create an efficient and compact deep learning model for crowd counting. The processing element, referred to as the shunting inhibitory neuron, generates complex decision boundaries, making it more powerful than the traditional perceptron. It is employed in both the encoder and decoder modules of the proposed model for feature extraction. Furthermore, the decoder includes alternating convolutional and transformer layers, which provide local receptive fields and global self-attention, respectively. This design captures rich contextual information that is used for generating accurate segmentation and density maps. The self-attention mechanism is implemented using convolution modulation instead of matrix multiplication to reduce computational costs. Experiments conducted on three challenging crowd counting datasets demonstrate that the proposed deep learning network, which comprises a small model size, achieves crowd counting performance comparable to that of state-of-the-art techniques. Codes are available at \u0000<uri>https://github.com/ftivive/SINet</uri>\u0000.","PeriodicalId":73305,"journal":{"name":"IEEE transactions on artificial intelligence","volume":"5 11","pages":"5733-5745"},"PeriodicalIF":0.0,"publicationDate":"2024-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142600171","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-13DOI: 10.1109/TAI.2024.3442153
Jun Fu;Yan Wang
Community detection is a fundamental and widely studied field in network science. To perform community detection, various competitive multiobjective evolutionary algorithms (MOEAs) have been proposed. It is worth noting that the latest continuous encoding (CE) method transforms the original discrete problem into a continuous one, which can achieve better community partitioning. However, the original CE ignored important structural features of nodes, such as the clustering coefficient (CC), resulting in poor initial solutions and reduced the performance of community detection. Therefore, we propose a simple scheme to effectively utilize node structure feature vectors to enhance community detection. Specifically, a CE and CC-based (CE-CC) MOEA called CECC-Net is proposed. In CECC-Net, the CC vector performs the Hadamard product with a continuous vector (i.e., a concatenation of the continuous variables $mathbf{x}$ associated with the edges), resulting in an improved initial individual. Then, applying the nonlinear transformation to the continuous-valued individual yields a discrete-valued community grouping solution. Furthermore, a corresponding adaptive operator is designed as an essential part of this scheme to mitigate the negative effects of feature vectors on population diversity. The effectiveness of the proposed scheme was validated through ablation and comparative experiments. Experimental results on synthetic and real-world networks demonstrate that the proposed algorithm has competitive performance in comparison with several state-of-the-art EA-based community detection algorithms.
社群检测是网络科学中的一个基础领域，也是一个被广泛研究的领域。为了进行社群检测，人们提出了各种有竞争力的多目标进化算法（MOEAs）。值得注意的是，最新的连续编码（CE）方法将原来的离散问题转化为连续问题，可以实现更好的社区划分。但是，原有的 CE 忽略了节点的重要结构特征，如聚类系数（CC），导致初始解不理想，降低了社区检测的性能。因此，我们提出了一种简单的方案，有效利用节点结构特征向量来增强社群检测。具体来说，我们提出了一种基于 CE 和 CC（CE-CC）的 MOEA，称为 CECC-Net。在 CECC-Net 中，CC 向量与连续向量（即与边缘相关的连续变量 $/mathbf{x}$）进行哈达玛乘积，从而得到一个改进的初始个体。然后，将非线性变换应用于连续值个体，就能得到离散值群体分组解决方案。此外，还设计了一个相应的自适应算子，作为该方案的重要组成部分，以减轻特征向量对群体多样性的负面影响。通过消融和对比实验，验证了所提方案的有效性。在合成网络和真实世界网络上的实验结果表明，与几种最先进的基于 EA 的群落检测算法相比，所提出的算法具有很强的竞争力。
{"title":"An Improved Continuous-Encoding-Based Multiobjective Evolutionary Algorithm for Community Detection in Complex Networks","authors":"Jun Fu;Yan Wang","doi":"10.1109/TAI.2024.3442153","DOIUrl":"https://doi.org/10.1109/TAI.2024.3442153","url":null,"abstract":"Community detection is a fundamental and widely studied field in network science. To perform community detection, various competitive multiobjective evolutionary algorithms (MOEAs) have been proposed. It is worth noting that the latest continuous encoding (CE) method transforms the original discrete problem into a continuous one, which can achieve better community partitioning. However, the original CE ignored important structural features of nodes, such as the clustering coefficient (CC), resulting in poor initial solutions and reduced the performance of community detection. Therefore, we propose a simple scheme to effectively utilize node structure feature vectors to enhance community detection. Specifically, a CE and CC-based (CE-CC) MOEA called CECC-Net is proposed. In CECC-Net, the CC vector performs the Hadamard product with a continuous vector (i.e., a concatenation of the continuous variables \u0000<inline-formula><tex-math>$mathbf{x}$</tex-math></inline-formula>\u0000 associated with the edges), resulting in an improved initial individual. Then, applying the nonlinear transformation to the continuous-valued individual yields a discrete-valued community grouping solution. Furthermore, a corresponding adaptive operator is designed as an essential part of this scheme to mitigate the negative effects of feature vectors on population diversity. The effectiveness of the proposed scheme was validated through ablation and comparative experiments. Experimental results on synthetic and real-world networks demonstrate that the proposed algorithm has competitive performance in comparison with several state-of-the-art EA-based community detection algorithms.","PeriodicalId":73305,"journal":{"name":"IEEE transactions on artificial intelligence","volume":"5 11","pages":"5815-5827"},"PeriodicalIF":0.0,"publicationDate":"2024-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142600419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This research intends to address the critical need for precise brain tumor prediction through the development of an automated method that entwines the Firefly (FF) algorithm and the interval type-II fuzzy (IT2FLS) technique. The proposed method improves tumor delineation in complex brain tissue by using the FF algorithm to find possible cluster positions and the IT2FLS system for final clustering. This algorithm demonstrates its versatility by processing diverse image sequences from BRATS challenge datasets (2017, 2018, and 2020), which encompass varying levels of complexity. Through comprehensive evaluation metrics such as sensitivity, specificity, and dice-overlap index (DOI), the proposed algorithm consistently yields improved segmentation results. Ultimately, this research aims to augment oncologists' perceptual acumen, facilitating enhanced intuition and comprehension of patients' conditions, thereby advancing decision-making capabilities in medical research.
{"title":"Intelligent Multigrade Brain Tumor Identification in MRI: A Metaheuristic-Based Uncertain Set Framework","authors":"Saravanan Alagarsamy;Vishnuvarthanan Govindaraj;A. Shahina;D. Nagarajan","doi":"10.1109/TAI.2024.3441520","DOIUrl":"https://doi.org/10.1109/TAI.2024.3441520","url":null,"abstract":"This research intends to address the critical need for precise brain tumor prediction through the development of an automated method that entwines the Firefly (FF) algorithm and the interval type-II fuzzy (IT2FLS) technique. The proposed method improves tumor delineation in complex brain tissue by using the FF algorithm to find possible cluster positions and the IT2FLS system for final clustering. This algorithm demonstrates its versatility by processing diverse image sequences from BRATS challenge datasets (2017, 2018, and 2020), which encompass varying levels of complexity. Through comprehensive evaluation metrics such as sensitivity, specificity, and dice-overlap index (DOI), the proposed algorithm consistently yields improved segmentation results. Ultimately, this research aims to augment oncologists' perceptual acumen, facilitating enhanced intuition and comprehension of patients' conditions, thereby advancing decision-making capabilities in medical research.","PeriodicalId":73305,"journal":{"name":"IEEE transactions on artificial intelligence","volume":"5 11","pages":"5381-5391"},"PeriodicalIF":0.0,"publicationDate":"2024-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142600363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-08DOI: 10.1109/TAI.2024.3440220
Linfang Yu;Zhen Qin;Liqun Xu;Zhiguang Qin;Kim-Kwang Raymond Choo
Human pose estimation (HPE) relies on the anatomical relationships among different body parts to locate keypoints. Despite the significant progress achieved by convolutional neural networks (CNN)-based models in HPE, they typically fail to explicitly learn the global dependencies among various body parts. To overcome this limitation, we propose a spatial-aware HPE model called SSpose that explicitly captures the spatial dependencies between specific key points and different locations in an image. The proposed SSpose model adopts a hybrid CNN-Transformer encoder to simultaneously capture local features and global dependencies. To better preserve image details, a multiscale fusion module is introduced to integrate coarse- and fine-grained image information. By establishing a connection with the activation maximization (AM) principle, the final attention layer of the Transformer aggregates contributions (i.e., attention scores) from all image positions and forms the maximum position in the heatmap, thereby achieving keypoint localization in the head structure. Additionally, to address the issue of visible information leakage in convolutional reconstruction, we have devised a self-supervised training framework for the SSpose model. This framework incorporates mask autoencoder (MAE) technology into SSpose models by utilizing masked convolution and hierarchical masking strategy, thereby facilitating efficient self-supervised learning. Extensive experiments demonstrate that SSpose performs exceptionally well in the pose estimation task. On the COCO val set, it achieves an AP and AR of 77.3% and 82.1%, respectively, while on the COCO test-dev set, the AP and AR are 76.4% and 81.5%. Moreover, the model exhibits strong generalization capabilities on MPII.
人体姿态估计（HPE）依赖于不同身体部位之间的解剖关系来定位关键点。尽管基于卷积神经网络（CNN）的模型在 HPE 方面取得了重大进展，但它们通常无法明确学习不同身体部位之间的全局依赖关系。为了克服这一局限，我们提出了一种名为 SSpose 的空间感知 HPE 模型，它能明确捕捉图像中特定关键点与不同位置之间的空间依赖关系。所提出的 SSpose 模型采用混合 CNN 变换器编码器，可同时捕捉局部特征和全局依赖关系。为了更好地保留图像细节，还引入了多尺度融合模块来整合粗粒度和细粒度图像信息。通过与激活最大化（AM）原理建立联系，变换器的最终注意力层汇总了来自所有图像位置的贡献（即注意力分数），并形成热图中的最大位置，从而实现头部结构中的关键点定位。此外，为了解决卷积重建中的可见信息泄漏问题，我们还为 SSpose 模型设计了一个自监督训练框架。该框架利用掩码卷积和分层掩码策略，将掩码自动编码器（MAE）技术融入 SSpose 模型，从而促进了高效的自我监督学习。大量实验证明，SSpose 在姿态估计任务中表现优异。在 COCO val 集上，它的 AP 和 AR 分别达到 77.3% 和 82.1%，而在 COCO test-dev 集上，AP 和 AR 分别为 76.4% 和 81.5%。此外，该模型在 MPII 上也表现出很强的泛化能力。
{"title":"SSpose: Self-Supervised Spatial-Aware Model for Human Pose Estimation","authors":"Linfang Yu;Zhen Qin;Liqun Xu;Zhiguang Qin;Kim-Kwang Raymond Choo","doi":"10.1109/TAI.2024.3440220","DOIUrl":"https://doi.org/10.1109/TAI.2024.3440220","url":null,"abstract":"Human pose estimation (HPE) relies on the anatomical relationships among different body parts to locate keypoints. Despite the significant progress achieved by convolutional neural networks (CNN)-based models in HPE, they typically fail to explicitly learn the global dependencies among various body parts. To overcome this limitation, we propose a spatial-aware HPE model called SSpose that explicitly captures the spatial dependencies between specific key points and different locations in an image. The proposed SSpose model adopts a hybrid CNN-Transformer encoder to simultaneously capture local features and global dependencies. To better preserve image details, a multiscale fusion module is introduced to integrate coarse- and fine-grained image information. By establishing a connection with the activation maximization (AM) principle, the final attention layer of the Transformer aggregates contributions (i.e., attention scores) from all image positions and forms the maximum position in the heatmap, thereby achieving keypoint localization in the head structure. Additionally, to address the issue of visible information leakage in convolutional reconstruction, we have devised a self-supervised training framework for the SSpose model. This framework incorporates mask autoencoder (MAE) technology into SSpose models by utilizing masked convolution and hierarchical masking strategy, thereby facilitating efficient self-supervised learning. Extensive experiments demonstrate that SSpose performs exceptionally well in the pose estimation task. On the COCO val set, it achieves an AP and AR of 77.3% and 82.1%, respectively, while on the COCO test-dev set, the AP and AR are 76.4% and 81.5%. Moreover, the model exhibits strong generalization capabilities on MPII.","PeriodicalId":73305,"journal":{"name":"IEEE transactions on artificial intelligence","volume":"5 11","pages":"5403-5417"},"PeriodicalIF":0.0,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142600097","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：