Submission Deadline: 31 December 2026 View: 105 Submit to Special Issue
Prof. Sidi Ahmed MAHMOUDI
Email: sidi.mahmoudi@umons.ac.be
Affiliation: Service of Computer Science, Software and Artificial Intelligence, Faculty of Engineering, University of Mons, Mons, Belgium
Research Interests: artificial intelligence, machine and deep learning, multimedia processing, XAI, multimodal learning
Dr. Mohamed BENKEDADRA
Email: mohamed.benkedadra@umons.ac.be
Affiliation: Service of Computer Science, Software and Artificial Intelligence, Faculty of Engineering, University of Mons, Mons, Belgium
Research Interests: deep learning, object detection, artificial intelligence
Dr. Otmane Amel
Email: otmane.amel@umons.ac.be
Affiliation: Service of Computer Science, Software and Artificial Intelligence, Faculty of Engineering, University of Mons, Mons, Belgium
Research Interests: artificial intelligence, deep learning, multimodal learning, computer networks and communications, control and systems engineering
In recent years, the integration of Artificial Intelligence (AI) with scientific computing, and high-performance computing (HPC) has significantly transformed engineering analysis and computational design. The convergence of Computer Vision, Natural Language, Big Data and Cloud Computing modeling has facilitated the emergence of hybrid modeling approaches that integrate data-driven learning with diverse data modalities.
Multimodal Artificial Intelligence has recently emerged as a promising paradigm for capturing complementary information from heterogeneous sources. By jointly processing images, textual data, temporal signals, and other structured or unstructured inputs, multimodal models can learn richer and more expressive representations of complex systems. Such capabilities have demonstrated substantial improvements in predictive performance, robustness, and generalization across a wide range of computational engineering applications. However, the increasing architectural complexity of multimodal learning systems introduces significant challenges related to interpretability, scalability, stability, and reliability. These issues are particularly critical in high-stakes engineering domains, where transparency, traceability, and trustworthiness are essential for ensuring safe and responsible deployment.
Although considerable progress has been made in the field of Explainable Artificial Intelligence (XAI), the majority of existing interpretability techniques have been primarily developed for unimodal learning systems, particularly those based on image data. Consequently, methodological frameworks for explaining and interpreting multimodal architectures remain relatively limited and insufficiently formalized. This lack of comprehensive interpretability approaches constitutes a major barrier to the broader adoption of multimodal AI in engineering contexts, where model decisions must often be understandable, verifiable, and aligned with domain knowledge.
This special issue aims to address these challenges by promoting novel research on explainable and interpretable multimodal AI methods specifically tailored to engineering design and scientific computing. Particular attention is given to advanced multimodal architectures relying on modular encoder-based representations combined with cross-modal fusion mechanisms. In addition, emerging vision–language models and other multimodal foundation models are considered as promising directions for improving robustness, transparency, and trustworthiness in complex computational pipelines.
In addition, this special issue explores the growing role of generative artificial intelligence in facilitating multimodal learning processes. Generative models can contribute to synthetic data generation, multimodal data augmentation, and cross-modal alignment, thereby alleviating data scarcity issues and improving training efficiency. These approaches offer new opportunities to enhance the adaptability and scalability of multimodal systems across diverse engineering environments.
By fostering interdisciplinary contributions at the intersection of scientific computing, high-performance computing, and multimodal artificial intelligence, this special issue seeks to advance the development of reliable, scalable, and interpretable intelligent systems for engineering computation and design. The proposed research directions are expected to impact a wide spectrum of application domains, including computer vision, Industry 4.0, medical imaging, multimedia information retrieval, fraud detection, software engineering, etc.
Research topics (not limited to):
This Special Issue welcomes original research contributions addressing theoretical, methodological, and applied aspects of multimodal artificial intelligence and explainable intelligent systems. Topics of interest include, but are not limited to:
· Multimodal learning and cross-modal representation learning
· Multimodal data fusion and heterogeneous data integration
· Multimodal foundation models and vision–language models
· Machine learning and deep learning for multimodal data analysis
· Advanced deep learning architectures (CNN, Transformer, ViT, LLM …)
· Explainable AI (XAI) and interpretable multimodal systems
· Trustworthy, transparent, and robust AI for engineering applications
· Generative AI, diffusion models, and synthetic data generation
· Self-supervised, unsupervised, and weakly supervised learning
· Neural Architecture Search (NAS) and automated ML (AutoML)
· Privacy-preserving AI and federated learning
· Natural Language Processing and multimodal understanding
· Multimedia information retrieval and cross-modal search
· Applications of multimodal AI in computer vision, Industry 4.0, medical imaging, fraud detection, software engineering, etc.
Login
Register
Submit a Paper
Propose a Special lssue