University of Pisa

Vision in Action: AI-Powered Systems for Sustainable Progress

A Special Session of the IEEE ICAI-TEMS 2025

This session explores how “Artificial Intelligence and Computer Vision” are enabling a wave of sustainable innovation — from ideation and prototyping to implementation, and deployment in the real world. As sustainable engineering practices become vital across industries, visual intelligence plays a central role in enhancing productivity, guiding decision-making, and managing complex systems in real-world environments. We encourage contributions that bridge the gap between research and practical application — whether through smart inspection systems in manufacturing, environmental monitoring via remote sensing, AI-powered healthcare technologies, or cross-sector collaborations that harmonize innovation with regulatory and societal needs. 


TOPICS:

From Research to Market: Vision-Based AI for Sustainable Systems
– From research to real-world deployment: Vision-based sustainability solutions
– Vision-based quality control, predictive maintenance, and automation
– Integrating AI into smart supply chains, circular economy models, and waste management systems
– Scalable/Efficient architectures for industrial edge-AI deployments

AI Integration for Operational Excellence and Sustainability
– Vision-powered process optimization in manufacturing and operations
– Remote sensing and digital twins for resource monitoring and land use planning
– AI for climate resilience and disaster response

Ethical and Societal Dimensions of Vision-Enabled Technologies
– Vision-based healthcare and assistive technologies (e.g., diagnostics, fall detection, mobility aids)
– Ethical, inclusive, and privacy-preserving AI applications in public infrastructure

Learning Methods, Data, and Transparency for practical AI use-cases 
– Deep learning and classical computer vision algorithms for quality control, monitoring, and automation
– Dataset contributions that enable practical deployment of AI across industrial and environmental domains
– Multi-modal learning approaches combining visual data with text, thermal, LiDAR, audio, or satellite inputs
– Unsupervised, weakly-supervised, semi-supervised, self-supervised, and continual learning methods for scalable and adaptive AI applications
– Explainable and transparent AI techniques for trustworthy decision-making in vision-based systems


ORGANIZERS & CONTACTS:
Inform us about your intention to submit, ask if you have any questions:

  • Mario G.C.A. Cimino, Dept of Information Engineering, University of Pisa, Italy, mario.cimino@unipi.it
  • Neelu Madan, Visual Analysis and Perception, Dept of Architecture, Design, and Media Technology, Aalborg University, Technical Faculty of IT and Design, Denmark, nema@create.aau.dk


DATES & SUBMISSION:
Papers will undergo a double-blind review process.

  • Submission deadline (to be extended).
  • Submission guidelines. After submission, the corresponding author is required to forward the submission notification email to mario.cimino@unipi.it and nema@create.aau.dk.