Common benchmarks help advance the entire field
Scientists are usually recognised for their own discoveries. The PAMI Mark Everingham Prize is different in this respect: it rewards work that enables others to do better research.
This is precisely where the significance of the Visual Object Tracking Challenge lies. Consider, for example, a system that needs to identify a car, a person or another object in a video and then reliably track it even as it changes position and appearance, moves quickly, becomes partially obscured or briefly disappears from view. Such technologies are used in fields including robotics, autonomous systems and automated video analysis.
To determine which method actually performs better and how much progress the field has made, however, researchers need common data, precisely defined tasks and standardised evaluation methods. VOT provides exactly this kind of infrastructure to the international research community.
Since 2013, VOT has been developing benchmarks, datasets, evaluation methodologies and open-source tools, enabling teams from around the world to test new algorithms under comparable conditions every year. The result is much more than a ranking of individual methods. Long-term, systematic comparisons help reveal their strengths and weaknesses, identify which research directions are genuinely driving progress, and formulate new problems for the field to address.
This contribution is captured in the citation on the award certificate. The team received the prize “for the series of Visual Object Tracking workshops and challenges that shaped the modern synthesis of visual object tracking.”
Professor Jiří Matas of the Department of Cybernetics at CTU FEE is among the researchers who have been involved with VOT since its beginnings. He works with colleagues from a number of European universities and research institutions on organising individual editions of the challenge, developing its methodology and producing its research outputs.
“Receiving an award as a team is a great joy because it is a joy we can share. At the same time, it has given us a strong boost at a moment when we really needed one. Organising an initiative like this involves a huge amount of work for which, in a sense, you get nothing in return, and after all these years we were beginning to run a little short of energy. The award has shown us that the community notices our work and considers it important. That has really given us renewed energy,” says Professor Jiří Matas, who has consistently been the highest-ranked computer scientist in the Czech Republic according to Research.com.
A second major international award within weeks
The PAMI Mark Everingham Prize is the second major international honour Jiří Matas has received within the space of just a few weeks. In August, he received the King-Sun Fu Prize at the ICPR 2026 conference in Lyon, France. The prize is the most prestigious award presented by the International Association for Pattern Recognition (IAPR). Matas received it for his key scientific contributions to geometric computer vision.
The two awards recognise different aspects of scientific work. While the King-Sun Fu Prize is an individual award honouring Matas’s scientific achievements, the Mark Everingham Prize highlights long-term teamwork carried out for the benefit of the wider research community.
At the same time, the two themes intersect in an interesting way. As the newly named recipient of the King-Sun Fu Prize, Matas opened ICPR with a lecture on benchmarking – the methods used to objectively measure and compare the quality of different solutions in computer vision. Establishing common benchmarks and conditions for comparing algorithms is also one of the fundamental principles behind VOT, the initiative that has now earned him and his colleagues another major distinction.
In the company of projects that transformed computer vision
The significance of the PAMI Mark Everingham Prize is also reflected in the list of its previous recipients. The award has previously gone to teams behind some of the most influential tools, datasets and benchmarks in modern computer vision, including ImageNet, OpenCV, COCO, KITTI, Caffe, Detectron and COLMAP. Its recipients also include researchers from some of the world’s leading universities and research institutions.
The prize commemorates British computer vision researcher Mark Everingham, who made major contributions to the development of the field, including through the creation of shared benchmarks and tools for the entire research community. Through the award, the IEEE Pattern Analysis and Machine Intelligence Technical Committee recognises precisely this kind of long-term activity whose results can be used by other researchers – from open datasets and software tools to challenges that enable standardised comparisons of methods.
For the VOT team, the award therefore represents not only recognition of more than a decade of work, but also encouragement to continue it. The initiative continues to evolve alongside the field itself, with its current challenges addressing increasingly complex tasks in object tracking and segmentation in video.
