CILab - Computational Intelligence Laboratory
Department of Computer Science, University of Bari Aldo Moro (Italy)
Welcome to the official Hugging Face organization of CILab. We are a research group dedicated to the advancement of Computational Intelligence, focusing on both fundamental theory and impactful real-world applications across diverse domains.
𧬠About Us
CILab is a research laboratory within the Department of Computer Science at the University of Bari Aldo Moro. Coordinated by Prof. Giovanna Castellano, our team works at the intersection of deep learning, fuzzy logic, and pattern recognition to build intelligent systems that are effective, interpretable, and human-centric.
π Research Pillars
Our research is structured around four main pillars:
- Digital Humanities & Multimedia: Leveraging AI for the analysis, clustering, and cross-modal generation of visual arts, music, and cultural heritage.
- Explainable AI (XAI) & Fuzzy Systems: Developing "interpretable-by-design" models, including fuzzy rule-based systems and linguistic summaries to explain black-box models.
- Computer Vision & Pattern Recognition: Specialized in semantic segmentation, few-shot learning, and lightweight architectures for aerial and drone imagery.
- Computational Intelligence in Healthcare: Applying AI to medical imaging, signal processing, and clinical decision support for neurodegenerative diseases and dysgraphia.
π Research Highlights
Our contributions span several innovative areas of machine learning:
- Cross-Modal & Generative AI: Research into bridging disparate modalities, such as connecting visual arts with music and using multimodal LLMs for guided image inpainting and textual report generation.
- Art Analysis & Knowledge Graphs: Combining deep learning with knowledge graphs for automatic artwork classification, link retrieval, and pattern extraction in paintings and drawings.
- Precision Agriculture & Environment: Developing lightweight vision transformers and density-based clustering for weed mapping and crowd detection from drone-captured imagery.
- Human-Centric Explainability: Creating semi-supervised fuzzy clustering and natural language explanations for complex data, ranging from acoustic signals to medical reports.
- Few-Shot & Prompt-Based Learning: Investigating multi-class semantic segmentation using visual prompts and few-shot paradigms to reduce data dependency.
π Connect With Us
βAdvancing intelligence through transparency and innovation.β