Complex & Intelligent Systems最新文献

{"title":"Rugularizing generalizable neural radiance field with limited-view images","authors":"Wei Sun, Ruijia Cui, Qianzhou Wang, Xianguang Kong, Yanning Zhang","doi":"10.1007/s40747-024-01696-6","DOIUrl":"https://doi.org/10.1007/s40747-024-01696-6","url":null,"abstract":"<p>We present a novel learning model with attention and prior guidance for view synthesis. In contrast to previous works that focus on optimizing for specific scenes with densely captured views, our model explores a generic deep neural framework to reconstruct radiance fields from a limited number of input views. To address challenges arising from under-constrained conditions, our approach employs cost volumes for geometry-aware scene reasoning, and integrates relevant knowledge from the ray-cast space and the surrounding-view space using an attention model. Additionally, a denoising diffusion model learns a prior over scene color, facilitating regularization of the training process and enabling high-quality radiance field reconstruction. Experimental results on diverse benchmark datasets demonstrate that our approach can generalize across scenes and produce realistic view synthesis results using only three input images, surpassing the performance of previous state-of-the-art methods. Moreover, our reconstructed radiance field can be effectively optimized by fine-tuning the target scene to achieve higher quality results with reduced optimization time. The code will be released at https://github.com/dsdefv/nerf.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"48 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142849200","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A robust adaptive meta-sample generation method for few-shot time series prediction","authors":"Chao Zhang, Defu Jiang, Kanghui Jiang, Jialin Yang, Yan Han, Ling Zhu, Libo Tao","doi":"10.1007/s40747-024-01638-2","DOIUrl":"https://doi.org/10.1007/s40747-024-01638-2","url":null,"abstract":"<p>The research and exploration of time series prediction (TSP) have attracted much attention recently. Researchers can achieve effective TSP based on the deep learning model and a large amount of data. However, when sufficient high-quality data are not available, the performance of prediction models based on deep learning techniques may degrade. Therefore, this paper focuses on few-shot time series prediction (FTSP) and plans to combine meta-learning and generative models to alleviate the problems caused by insufficient training data. When using meta-learning techniques to process FTSP tasks, researchers set the meta-parameter in model-agnostic meta-learning (MAML) as a meta-sample and construct meta-sample generation methods based on advanced generative modeling theory to achieve better uncertainty coding. The existing meta-sample generation methods in FTSP scenes have an inherent limitation: With the increase of the complexity of prediction tasks, samples based on Gaussian distribution may be sensitive to noise and outliers in the meta-learning environment and lack of uncertainty expression, thus affecting the robustness and accuracy of prediction. Therefore, this paper proposes an adaptive sample generation method called JLSG-Diffusion. Based on the Jensen constraint framework and Laplace modeling theory, this method constructs a sample adapter with reasonable adaptive steps and fast convergence for specific tasks. The advantage is to realize fast adaptive convergence of samples to new tasks at lower cost, effectively control the overall generalization error, and improve the robustness and non-Gaussian generalization of sample posterior reasoning. Moreover, the meta sampler of JLSG-Diffusion embeds meta-learning from the implicit probability measure level of Denoising Diffusion Probabilistic Models (DDPM), which makes the meta-sample distribution directly establish a function mapping with the new task and effectively quantifies the uncertainty of spatiotemporal dimension. Experimental results on three real datasets prove the efficiency and effectiveness of JLSG-Diffusion. Compared with the benchmark methods, the prediction model combined with JLSG-Diffusion shows better accuracy.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"23 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142848868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Division-selection transfer learning for prediction based dynamic multi-objective optimization","authors":"Hongye Li, Fan Liang, Yulu Liu, Quanheng Zheng, Kunru Guo","doi":"10.1007/s40747-024-01656-0","DOIUrl":"https://doi.org/10.1007/s40747-024-01656-0","url":null,"abstract":"<p>Dynamic multi-objective optimization problems (DMOPs) are challenging as they require capturing the Pareto optimal front (POF) and Pareto optimal set (POS) during the optimization process. In recent years, transfer learning (TL) has emerged the empirical knowledge and is an effective approach to solve DMOPs. However, negative transfer can occur when the transfer method is not suitable for the transfer task. It may deviate the search path and seriously reduce the efficiency, so how to reduce the occurrence of negative transfer to save the running time of dynamic multi-objective evolutionary algorithm (DMOEA) is an important issue to be addressed. A division-selection transfer learning evolutionary algorithm for dynamic multi-objective optimization (DST-DMOEA) is designed towards this aim. Specifically, individuals with high Spearman correlation are relatively stable in different environments, selecting them to train the Support Vector Regression (SVR) model ensures a more accurate capture of solution features, predicting the objective values of historical solutions based on the model, and thus divide historical solutions into elite and non-elite solutions. Subsequently, for the elite solutions, individual TL that incorporates local information for optimization and transfer is used, while the non-elite solutions are handled with manifold TL method to obtain the overall data distribution and understand the internal structure. Then, merge the predicted individuals generated by two parts of TL will constitute as the initial population in the optimization process. Compared with other algorithms, the initial solution of DST-DMOEA is closer to the real POF, effectively reducing negative transfer. In addition, in 51 test instances, DST-DMOEA has shown superior performance in over 30 instances.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"88 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142849201","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MCPA: multi-scale cross perceptron attention network for 2D medical image segmentation","authors":"Liang Xu, Mingxiao Chen, Yi Cheng, Pengwu Song, Pengfei Shao, Shuwei Shen, Peng Yao, Ronald X. Xu","doi":"10.1007/s40747-024-01671-1","DOIUrl":"https://doi.org/10.1007/s40747-024-01671-1","url":null,"abstract":"<p>The UNet architecture, based on convolutional neural networks (CNN), has demonstrated its remarkable performance in medical image analysis. However, it faces challenges in capturing long-range dependencies due to the limited receptive fields and inherent bias of convolutional operations. Recently, numerous transformer-based techniques have been incorporated into the UNet architecture to overcome this limitation by effectively capturing global feature correlations. However, the integration of the Transformer modules may result in the loss of local contextual information during the global feature fusion process. In this work, we propose a 2D medical image segmentation model called multi-scale cross perceptron attention network (MCPA). The MCPA consists of three main components: an encoder, a decoder, and a Cross Perceptron. The Cross Perceptron first captures the local correlations using multiple Multi-scale Cross Perceptron modules, facilitating the fusion of features across scales. The resulting multi-scale feature vectors are then spatially unfolded, concatenated, and fed through a Global Perceptron module to model global dependencies. Considering the high computational cost of using 3D neural network models, and the fact that many important clinical data can only be obtained in two dimensions, our MCPA focuses on 2D medical image segmentation. Furthermore, we introduce a progressive dual-branch structure (PDBS) to address the semantic segmentation of the image involving finer tissue structures. This structure gradually shifts the segmentation focus of MCPA network training from large-scale structural features to more sophisticated pixel-level features. We evaluate our proposed MCPA model on several publicly available medical image datasets from different tasks and devices, including the open large-scale dataset of CT (Synapse), MRI (ACDC), and widely used 2D medical imaging datasets captured by fundus camera (DRIVE, CHASE<span>(_)</span>DB1, HRF), and OCTA (ROSE). The experimental results show that our MCPA model achieves state-of-the-art performance.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"23 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142849198","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dynamic decomposition and hyper-distance based many-objective evolutionary algorithm","authors":"Xujian Wang, Fenggan Zhang, Minli Yao","doi":"10.1007/s40747-024-01637-3","DOIUrl":"https://doi.org/10.1007/s40747-024-01637-3","url":null,"abstract":"<p>Nowadays many algorithms have appeared to solve many-objective optimization problems (MaOPs), yet the balance between convergence and diversity is still an open issue. In this paper, we propose a dynamic decomposition and hyper-distance based many-objective evolutionary algorithm named DHEA. On one hand, to maximize the diversity of the population, we use dynamic decomposition to decompose the whole population into multiple clusters. Specifically, first find pivot solutions according to the distribution of the population through the max–min-angle strategy, and then, assign solutions into different clusters according to their distances to pivot solutions. On the other hand, to select solutions from each cluster with balanced convergence and diversity, we propose hyper-distance based angle penalized distance for fitness assignment. Specifically, first compute the distance of solutions to the hyperplane and to the pivot solution to measure convergence and diversity, respectively, and then select the solution with the smallest fitness value. Hyper-distance, as convergence-related component, alleviates the bias towards problems with concave PFs. Besides, to promote convergence, the concept of knee points is introduced to mating selection. Through comparison with nine algorithms on 27 test problems, DHEA is validated to be effective and competitive to deal with MaOPs with different types of Pareto fronts and stable on different numbers of objectives.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"30 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142848867","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The algorithm for foggy weather target detection based on YOLOv5 in complex scenes","authors":"Zhaohui Liu, Wenshuai Hou, Wenjing Chen, Jiaxiu Chang","doi":"10.1007/s40747-024-01679-7","DOIUrl":"https://doi.org/10.1007/s40747-024-01679-7","url":null,"abstract":"<p>With the rapid development of urbanization and global climate warming, complex environments and adverse weather conditions pose significant challenges to the accuracy of object detection and driving safety in autonomous driving systems. Addressing the challenges of severe occlusion and noise interference in target detection within complex foggy scenes and considering the YOLOv5 model’s small size and fast processing capabilities, which meet the real-time processing demands in complex environments, it is particularly suited for resource-constrained vehicular systems. Consequently, this paper introduces the YOLOv5-RCBiW model tailored for vehicular vision perception aimed at enhancing feature extraction and recognition. Initially, the Receptive Field Block (RFB) is integrated with the Coordinate Attention (CA) mechanism to form the RFCA module, which emphasizes the importance of different features and optimizes receptive field spatial features. Furthermore, the Re-BiFPN module is constructed to enhance feature perception accuracy through bidirectional cross-scale connections and feature fusion, while the detection head at the P5 layer is replaced to improve recognition capabilities. Finally, a gradient gain loss function is introduced to reduce feature information loss and prevent model performance degradation, ensuring robustness and accuracy in complex environments. The comparative experimental results on the RTTS and Foggy Driving datasets indicate that the YOLOv5-RCBiW model significantly outperforms existing models in object detection accuracy under foggy and complex scenes. Additionally, in-vehicle experiments validate the model’s effectiveness and real-time performance in challenging environments.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"115 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142849192","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Tactical intent-driven autonomous air combat behavior generation method","authors":"Xingyu Wang, Zhen Yang, Shiyuan Chai, Jichuan Huang, Yupeng He, Deyun Zhou","doi":"10.1007/s40747-024-01685-9","DOIUrl":"https://doi.org/10.1007/s40747-024-01685-9","url":null,"abstract":"<p>With the rapid development and deep application of artificial intelligence, modern air combat is incrementally evolving towards intelligent combat. Although deep reinforcement learning algorithms have contributed to dramatic advances in in air combat, they still face challenges such as poor interpretability and weak transferability of adversarial strategies. In this regard, this paper proposes a tactical intent-driven method for autonomous air combat behaviour generation. Firstly, this paper explores the mapping relationship between optimal strategies and rewards, demonstrating the detrimental effects of the combination of sparse rewards and dense rewards on policy. Built around this, the decision-making process of pilot behavior is analyzed, and a reward mapping model from intent to behavior is established. Finally, to address the problems of poor stability and slow convergence speed of deep reinforcement learning algorithms in large-scale state-action spaces, the dueling-noisy-multi-step DQN algorithm is devised, which not only improves the accuracy of value function approximation but also enhances the efficiency of space exploration and network generalization. Through experiments, the conflicts between sparse rewards and dense rewards are demonstrated. The superior performance and stability of the proposed algorithm compared to other algorithms are captured by our empirical results. More intuitively, the strategies under different intents exhibit strong interpretability and flexibility, which can provide tactical support for intelligent decision-making in air combat.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"262 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142776652","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A survey of MRI-based brain tissue segmentation using deep learning","authors":"Liang Wu, Shirui Wang, Jun Liu, Lixia Hou, Na Li, Fei Su, Xi Yang, Weizhao Lu, Jianfeng Qiu, Ming Zhang, Li Song","doi":"10.1007/s40747-024-01639-1","DOIUrl":"https://doi.org/10.1007/s40747-024-01639-1","url":null,"abstract":"<p>Segmentation of brain tissue from MR images provides detailed quantitative brain analysis for accurate diagnosis, detection, and classification of brain diseases, and plays an important role in neuroimaging research and clinical environments. Recently, a plethora of deep learning-based approaches have been employed to achieve brain tissue segmentation in fetuses, infants, and adults with impressive outcomes. However, owing to the existence of noise, motion artifacts, and edge blurriness in MR images, automatically segmenting brain tissue accurately from MR images is still a very challenging task. This survey examines both deep learning and MRI, providing an overview of the latest advances in fetal, infant, and adult brain tissue segmentation techniques based on deep learning. It includes the performance and quantitative analysis of the state-of-the-art methods. Over 100 scientific papers covering various technical aspects, including network architecture, prior knowledge, and attention mechanisms, were reviewed and analyzed. This article also comprehensively discusses these technologies and their potential applications in the future. Brain tissue segmentation provides detailed quantitative brain analysis for accurate diagnosis, detection, and classification of brain diseases, and plays an important role in neuroimaging research and clinical environments.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"31 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142776603","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated generation of dispatching rules for the green unrelated machines scheduling problem","authors":"Nikolina Frid, Marko Ɖurasević, Francisco Javier Gil-Gala","doi":"10.1007/s40747-024-01677-9","DOIUrl":"https://doi.org/10.1007/s40747-024-01677-9","url":null,"abstract":"<p>The concept of green scheduling, which deals with the environmental impact of the scheduling process, is becoming increasingly important due to growing environmental concerns. Most green scheduling problem variants focus on modelling the energy consumption during the execution of the schedule. However, the dynamic unrelated machines environment is rarely considered, mainly because it is difficult to manually design simple heuristics, called dispatching rules (DRs), which are suitable for solving dynamic, non-standard scheduling problems. Using hyperheuristics, especially genetic programming (GP), alleviates the problem since it enables the automatic design of new DRs. In this study, we apply GP to automatically design DRs for solving the green scheduling problem in the unrelated machines environment under dynamic conditions. The total energy consumed during the system execution is optimised along with two standard scheduling criteria. The three most commonly investigated green scheduling problem variants from the literature are selected, and GP is adapted to generate appropriate DRs for each. The experiments show that GP-generated DRs efficiently solve the problem under dynamic conditions, providing a trade-off between optimising standard and energy-related criteria.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"219 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142776653","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DVGEDR: a drug repositioning method based on dual-view fusion and graph enhancement mechanism in heterogeneous networks","authors":"Dongjiang Niu, Lianwei Zhang, Beiyi Zhang, Qiang Zhang, Shanyang Ding, Hai Wei, Zhen Li","doi":"10.1007/s40747-024-01674-y","DOIUrl":"https://doi.org/10.1007/s40747-024-01674-y","url":null,"abstract":"<p>Drug repositioning, the discovery of new therapeutic uses for existing drugs, is increasingly gaining attention as a cost-effective and high-yield drug discovery strategy. Existing methods integrate diverse biological information into heterogeneous networks, providing a comprehensive framework for understanding complex drug–disease associations, which also will introduces noise into the data and affect the performance of the model. In this paper, first, a novel drug repositioning method, namely DVGEDR, is proposed, which generates two subgraphs of the target drug–disease pair to fuse biological information and integrate drug–disease associations from two distinct perspectives: drug–disease heterogeneous network and similarity networks. Next, a Multiple Attention Graph encoder (MAGencoder) module is designed to learn subgraph features and explore relationships between entities, which also improve the interpretability of the model. Finally, a graph enhancement mechanism is devised to improve the perception of critical information of model, enabling the model to flexibly process different graph structures. Performance comparisons with baseline models on three public datasets validate the state-of-the-art performance of DVGEDR in the field of drug repositioning. In case study, DVGEDR identifies 10 new candidate drugs for breast cancer and COVID-19, demonstrating not only superior performance in experimental settings but also potential therapeutic advantages in clinical environments. Furthermore, we select two sets of instances and further analyzed the attention distribution of the different nodes in the subgraph to explain the decision process of the model.</p>","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"19 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2024-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142776654","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}