Prof. Dr. Jiri Matas | AI Symposium 2026 | Plenary Talk скачать в хорошем качестве

Prof. Dr. Jiri Matas | AI Symposium 2026 | Plenary Talk

We begin the second day of AI Symposium 2026 with a plenary talk by Prof. Dr. Jiri Matas, Head of the Visual Recognition Group at the Czech Technical University in Prague. A leading figure in visual recognition, object detection, image retrieval and pattern recognition, Prof. Matas has helped shape the foundations of modern computer vision through more than 300 scientific publications and decades of influential research and academic leadership. In his plenary talk, “Benchmarking for Visual Recognition in the Era of VLMs and LLMs,” he examines one of the central challenges of today’s AI landscape: how to reliably measure progress in systems trained on internet-scale data. From the limitations of widely used benchmarks such as ImageNet to emerging questions around fair evaluation and data overlap, his talk highlights the need for more rigorous frameworks to assess real advances in visual intelligence.