AIDA4Edge

Short-Term Visit of the University of Niš Team to the University of Manchester

—

by

aida4edge
in

From September 23–26, 2025, the University of Niš team participated in a short-term visit hosted by the University of Manchester, Department of Computer Science, as part of the EU- and UKRI-funded Twinning AIDA4Edge project. This visit brought together experienced researchers and early stage researchers from both sides, as well as researchers from other universities and research centers across Europe, to exchange knowledge and strengthen collaboration in the fields of dynamic neural networks and spiking neural networks. The program also included sessions on research management and administration, covering topics such as gender equality, diversity, and inclusion within the research environment.

Over the course of four days, we had the opportunity to engage in a rich and inspiring program featuring cutting-edge research, hands-on training, and meaningful discussions:

🔹 September 23 – Dynamic Neural Networks Day Short-Term Visit of the University of Niš Team to the University of Manchester

  • Prof. Davide Bertozzi(University of Manchester): “Dynamic Neural Networks – a Review” (Parts I & II)
  • Meeting with Octavian Buiu (Scientific Director of the National R&D Institute for Microtechnologies _ IMT Bucharest) and Marius Mitroi form UEFISCDI (Executive Agency for Higher Education, Research, Development and Innovative Funding)
  • Online Mini-Workshop on Dynamic Neural Networks (Session I: Network Architecture Search and Session II: Slimmable Neural Networks):
    • Prof. Riccardo Zese (University of Ferrara, Italy): “Efficient Resource-Aware Neural Architecture Search with a Neuro-Symbolic Approach”
    • Matteo Gambella (Politecnico di Milano, Italy): “NACHOS: Neural Architecture Search for Hardware Constrained Early Exit Neural Networks”
    • Madhi Taheri (Taltech, Estonia & BTU Cottbus-Senftenberg, Germany): “RL-Agent-based Early-Exit DNN Architecture Search Framework”
    • Modhe Almelihi (University of Manchester, UK): “Design Strategies for Adaptive Execution in Slimmable Neural Architectures”
    • Rama Mounika Kodamanchili (Taltech, Estonia): “Adaptive Fault Resilience for Early-Exit DNNs”
    • Michalis Piponidis (University of Cyprus, Cyprus): “Dynamic Early-Exit Convolutional Neural Networks for Edge Vision: The Benefits, The Challenges, and the Road Ahead”

🔹 September 24 – Spiking Neural Network Day

  • Edward Jones (University of Manchester): “Biologically-Inspired Learning with SNNs” (lectures and hands-on sessions)
  • Edward Jones (University of Manchester): “Machine Learning with SNNs” (lectures and hands-on sessions)

🔹September 25 – Neuromorphic Vision Day

  • Computer Science Department Seminar

Guest lecture – Prof. Jelena Nikolić: “Quantization and Pruning: Key Techniques for Neural Network Compression”

  • AIDA4Edge short research-oriented presentations:
    • Prof. Aleksandra Jovanović: “An Approach to Weight Representation in Artificial and Spiking Neural Networks”
    • Junior Researcher Sofija Perić: “SQNR and Experimental Analysis of the Power-of-Two 3-Bit Quantizer for Neural Network Weight Quantization”
    • Prof. Dejan Ćirić: “Case Studies on the Application of Supervised and Unsupervised Learning to Acoustic Data Analysis”
    • Ph.D. student Tamara Milovanović: “Dynamic by Design: Shaping the Next Generation of Edge AI”
  • Edward Jones: “Seeing with Events: A Practical Introduction to Processing Visual Events” (hands-on training with event-based cameras)

🔹 September 26

  • Rachel Cowen and Vidyadhar Peesapati (University of Manchester):“EDI Strategy and Athena Swan Progress and Actions”

This short-term visit highlighted the importance of cross-institutional collaborationand provided valuable opportunities for knowledge transfer, training, and joint exploration of innovative AI methodologies.

We look forward to continuing this successful collaboration within the framework of the AIDA4Edge project.