Complex & Intelligent Systems最新文献

Semantic generalisable multimodal AI generated content detection via mixture of orthogonal decomposition and product manifold disentanglement 基于正交分解和积流形解纠缠的语义泛化多模态人工智能生成内容检测

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-20 DOI: 10.1007/s40747-026-02536-5

Qian Liu, Sonja Schimmler, Manfred Hauswirth

{"title":"Semantic generalisable multimodal AI generated content detection via mixture of orthogonal decomposition and product manifold disentanglement","authors":"Qian Liu, Sonja Schimmler, Manfred Hauswirth","doi":"10.1007/s40747-026-02536-5","DOIUrl":"https://doi.org/10.1007/s40747-026-02536-5","url":null,"abstract":"The generalisability of AI-Generated Content (AIGC, e.g., images and videos) detection can be decomposed into two tightly coupled subspaces: a semantic subspace and a forgery subspace. State-of-the-art AIGC detectors primarily focus on enhancing generalisability within the forgery subspace, typically neglecting the influences of the semantic subspace. Moreover, most existing AIGC detection methods consolidate all forged images and videos produced by heterogeneous deepfake generation techniques into a single “fake” class. This design choice overlooks the distinctive characteristics and substantial variability associated with different deepfake generation techniques. To address these limitations, we propose a novel Semantic Generalisable Deepfake Detection (SGDFD) framework that integrates a sparsely gated Mixture of ORthogonal Decomposition (MoORD) architecture. The parameter-efficient fine-tuning strategy based on orthogonal decomposition enhances the generalisation capability of the SGDFD model in AIGC detection, while the MoORD architecture further improves its semantic generalisability. In addition, the SGDFD model learns a product-manifold representation that decomposes the monolithic “fake” class into multiple sub-classes, each corresponding to a distinct deepfake generation method. Extensive experiments conducted on 10 real-world AIGC detection datasets under 3 evaluation protocols demonstrate that the proposed SGDFD model achieves state-of-the-art generalisation performance in AIGC detection across both known and unknown semantic categories.","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"65 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2026-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149221441","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}

引用次数: 0

Optimized multi-scale dual-modality feature pyramid network for early eye disease classification using fundus and OCT imaging 基于眼底和OCT影像的多尺度双模态特征金字塔网络优化早期眼病分类

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-20 DOI: 10.1007/s40747-026-02507-w

R. Kiruthiga, A. Saran Kumar

引用次数: 0

Explainable topology-aware frequency-semantic transformer for robust multi-class retinal OCT diagnosis 用于多类别视网膜OCT诊断的可解释的拓扑感知频率语义转换器

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-18 DOI: 10.1007/s40747-026-02571-2

Jeena Joseph

引用次数: 0

Drift-aware and privacy-preserving explainable credit scoring for fair decision-making in real world deployment 现实世界部署中公平决策的漂移感知和保护隐私的可解释信用评分

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-17 DOI: 10.1007/s40747-026-02530-x

Muralikrishnan Mani,Sonaa Rajagopal,K. Hemalatha

引用次数: 0

AI-driven whole-life costing for decision-ready UK public building retrofits 人工智能驱动的英国公共建筑改造决策的全寿命成本计算

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-16 DOI: 10.1007/s40747-026-02516-9

Ziru Tao,Paul Baguley,Rashid Maqbool,Obuks Ejohwomu

{"title":"AI-driven whole-life costing for decision-ready UK public building retrofits","authors":"Ziru Tao,Paul Baguley,Rashid Maqbool,Obuks Ejohwomu","doi":"10.1007/s40747-026-02516-9","DOIUrl":"https://doi.org/10.1007/s40747-026-02516-9","url":null,"abstract":"Abstract Public-sector building retrofits in the United Kingdom require decisions that balance whole-life cost, carbon reduction, service continuity and auditability under limited data and constrained capital budgets. This paper develops a decision-ready whole-life costing (WLC) workflow that integrates environmental, social and governance (ESG) priorities with uncertainty-aware artificial intelligence (AI) for public-building retrofit appraisal. The workflow combines national non-domestic Energy Performance Certificate (EPC) and Display Energy Certificate (DEC) data, Public Sector Decarbonisation Scheme (PSDS) delivery evidence and stakeholder inputs to construct a scenario library of public-building archetypes, retrofit actions and appraisal constraints. Stakeholder priorities from interviews, a questionnaire and a co-design workshop are translated into transparent weights using fuzzy analytic hierarchy process (F-AHP). These weights parameterise a reinforcement learning (RL) planner, while a Bayesian convolutional neural network–long short-term memory (BCNN–LSTM) surrogate provides uncertainty-bounded energy, carbon and cost trajectories for repeated policy evaluation. Results show that the workflow supports auditable retrofit sequencing across archetype and budget conditions and improves composite ESG performance relative to conventional WLC-only, £/tCO 2 e and static multi-criteria decision analysis (MCDA) baselines. For 1000 matched archetype–budget cases, the trained deep Q-network (DQN) decision-search stage achieves a 30.03-fold speedup over traditional scalar feasible-sequence enumeration. The contribution is an integrated and explainable WLC–ESG–AI workflow for early-stage public retrofit decision support. The current validation remains based on public datasets and scenario-level evidence so future work should extend the framework using project-specific metered data, digital twins and live portfolio deployment.","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"41 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2026-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149209359","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}

引用次数: 0

CHM-DTKT: a causal hypergraph selective state space digital twin for deconfounded and temporally grounded knowledge tracing CHM-DTKT：一个因果超图选择状态空间数字孪生，用于去基础和时间基础知识跟踪

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-16 DOI: 10.1007/s40747-026-02558-z

Qian Chen

引用次数: 0

Hybrid transformer-based recommender system with LLM-assisted semantic modeling for sequential and federated learning 基于混合变压器的推荐系统与llm辅助语义建模的顺序和联邦学习

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-14 DOI: 10.1007/s40747-026-02529-4

Maddala Lakshmi Bai,Rajendra Pamula,Subbarao Katteda

引用次数: 0

See other, focus self: local-to-global cross-modal guidance for infrared and visible image fusion 参见other， focus self: local-to-global交叉模态制导，用于红外和可见光图像融合

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-14 DOI: 10.1007/s40747-026-02547-2

Shuang Ma,Jiangbin Zheng,Yang Li,Chengcheng Ma

引用次数: 0

Anchor placement optimization in outdoor UWB systems for pedestrian-surrogate localization 户外超宽带系统中行人替代定位的锚点放置优化

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-13 DOI: 10.1007/s40747-026-02542-7

Byeonghyun Lee,Jeik Choi,JungHun Choi,Ki-Yong Oh

引用次数: 0

Carbon price forecasting based on time-lagged cross-correlation dynamic graphs and a dual-stream spatiotemporal fusion network 基于滞后互相关动态图和双流时空融合网络的碳价预测

IF 5.8 2区计算机科学

Complex & Intelligent Systems Pub Date : 2026-10-13 DOI: 10.1007/s40747-026-02550-7

Pingrang Wang,Xi Qiao,Haiyang Liu,Xiang Wang

引用次数: 0