One of the major hurdles on the road towards automation and AI is the lack of labelled training data in a variety of important, real-world domains. This is a well-known problem for practitioners who aim to apply state-of-the-art AI techniques on industrial automation problems. Data sparsity keeps this set of powerful technologies out of reach of most industry participants (apart from the very few industry giants) as well as most public sector bodies.

 

Generating synthetic training data has recently become a viable alternative where we train a high-level generative model whose labelled output will be used downstream in other AI architectures. In this talk, I will use a set of recent case studies to try to infer some basic principles about this paradigm. Can we predict when will this approach work and when it will not? What are the suitable domains and what are the necessary ingredients in our models? While a full theory is still not available, I hope to be able to demonstrate this is an interesting problem to study, as well as having huge potential impact.

After obtaining his MSci in Mathematics and Computer Science from Imperial College, London, George Vogiatzis embarked on PhD studies in Computer Vision from Trinity College, Cambridge in 2006 working under Prof Roberto Cipolla and Prof Philip Torr in the Machine Intelligence Laboratory of the University of Cambridge. He then spent three years as a Research Scientist in Toshiba Research Cambridge, concurrently holding a Junior Research Fellowship at Wolfson College, Cambridge before joining Aston University in 2010 as a Lecturer. He became a member of the CS department at Loughborough in 2023. He is an expert in 3D Computer Vision, and deep learning, with an emphasis in recent years on the synthetic generation of training datasets for domains where data is sparse.

 

He publishes regularly in top-tier journals and conferences in Computer Vision and has an extensive network of industrial collaborators. GV has secured over £2.5M of funding from industry, UKRI and EU sources..

 

---

 

 
About the EuroCC-2 project

EuroCC 2 will work to identify and address the skills gaps in the European High Performance Computing (HPC) ecosystem and coordinate cooperation across Europe to ensure a consistent skills base.

The role of EuroCC 2 is to establish and run a network of more than 30 NCCs across the EuroHPC Participating States. The NCCs act as single points of access in each country between stakeholders and national and EuroHPC systems. They operate on a regional and national level to liaise with local communities, in particular SMEs, map HPC competencies and facilitate access to European HPC resources for users from the private and public sector.

EuroCC 2 delivers training, interacts with industry, develops competence mapping and communication materials and activities, and supports the adoption of HPC services in other related fields, such as quantum computing, artificial intelligence (AI), high performance data analytics (HPDA) to expand the HPC user base.

 

---

 

This project has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 101101903. The JU receives support from the European Union’s programme Digital Europe and Germany, Bulgaria, Austria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, Greece, Hungary, Ireland, Italy, Lithuania, Latvia, Poland, Portugal, Romania, Slovenia, Spain, Sweden, France, Netherlands, Belgium, Luxembourg , Slovakia, Norway, Switzerland, Turkey, the Republic of North Macedonia, Iceland, Montenegro and Serbia. 

 

---