Edinburgh ranks among the world’s top centres for Computer Science and Engineering, a status that is underscored by Google’s research awards. Image Left to right: Vittorio Ferrari, Jakub Konecný and Christophe Dubach Jakub Konecný, one of three people at the University who have received a Google European Doctoral Fellowship, worth $180,000 over three years, is immersed in Optimization Algorithms. The 23-year-old Slovakian is busy solving such problems as how to generate recommendations on YouTube. “The area I’m working in is something between mathematics, statistics and computer science,” he says. “It is strongly related to Machine Learning, which is part of the area of artificial intelligence – which, in short, tries to describe how computers should learn new things based on available data. “I had a good idea of what I wanted to do – not just Machine Learning but the mathematics behind it – and was lucky enough that one of the leading experts in this field, Peter Richtárik, was here at Edinburgh University.” Typical Machine Learning, he explains, is about how you model a real problem in mathematical terms, which usually ends up as an optimization problem. “The goal is to find the parameters within which a learning model works best. I’m working on this, especially how to do so in a large-scale setting. For instance, with YouTube recommendations, the function is described in terms of a huge table, where some columns represent YouTube users, and others represent videos – it’s roughly a billion times a billion. “This is very large problem that needs to be solved every time you want to see a video.” Jakub can draw on Archer (Academic Research Computing High End Resource), the UK’s biggest supercomputer, which was launched at the University earlier in 2014 and holds more than 70,000 cores. I had a good idea of what I wanted to do – not just Machine Learning but the mathematics behind it – and was lucky enough that one of the leading experts in this field, Peter Richtárik, was here at Edinburgh University. Jakub Konecný Google has also awarded four Faculty Research Awards to Edinburgh. One is to Vittorio Ferrari, who is originally from Switzerland and is a Reader at the University’s School of Informatics. He also leads the CALVIN research group on visual learning. Vittorio received $73,000 for his project “Learning physical part models from YouTube videos”. The model is used on any given new image to answer questions such as: is there a tiger (or cat, or person etc.) in this image? Where is it exactly? What is its body pose? Which action is it performing? A computer vision program first needs to be trained for this task. Traditionally, this has been done with photographs but that is difficult as they do not show a range of articulations. “You need many photographs and many correspondences between points in different photographs,” Ferrari says. “There are a number of applications for visual recognition. For instance you could determine the spatial configuration of a person’s body parts or you could use it to animate an avatar, as in computer graphics in movies, where you see some creature or animal mimicking the movements of the actor. “Or you might want to find all the images of your cat jumping from the sofa, or all the scenes where James Bond is shooting someone.” The Google Faculty Research Award enables Vittorio to extend this line of work towards large-scale scenarios, he says. “Being in contact with the company also gives me access to their wonderful facilities. “The University is the top place in Europe for informatics, especially the artificial intelligence group of people,” he adds. David Harper, Head of EMEA University Relations at Google, says the award schemes help build and sustain strong relationships between the company and the academic community, especially in computer science, computer engineering and related fields. Jakub can draw on Archer (Academic Research Computing High End Resource), the UK’s biggest supercomputer, which was launched at the University earlier in 2014 and holds more than 70,000 cores. “Google was born in Stanford’s Computer Science department and the company has always maintained and valued strong relations with universities and research institutes. “We run a variety of programmes that provide funding and resources to the academic and external research community. Google, in turn, learns from the community through exposure to a broader set of ideas and approaches.” Google Faculty Research Awards are structured as unrestricted gifts to the University. “In the best cases, the transfer of know-how from university to Google, and vice versa, can transform both research and practice in the field of computer science,” David says. How does Google rate Edinburgh as a global centre for Computer Science and Engineering? “It clearly ranks highly,” David says. “In the March 2014 round, the University of Edinburgh received four of our Faculty Research Awards. Only four other universities around the world secured at least that many, and they were all in the US.” Another of those recipients is Christophe Dubach, also from Switzerland, who did a PhD in Informatics at the University after graduating from EPFL in Lausanne. “I’m interested in how we can make computer programs more efficient both in terms of performance and energy. “Energy is now driving the design of new hardware and, as a result, a wide range of specialised devices have appeared. Using cars as an analogy, some might be great for competing on a racing track while others might be best suited for the city. The real question is how to create programming abstraction that can be used to write programs in a hardware-agnostic way,” Christophe says. “One way to make this happen in the software world is to look at algorithmic patterns, which basically help programmers express their problems independently of the hardware.” Christophe says the Google Faculty Research Award (worth $53,573) will allow him to widen his research in this direction and support more members in his team. “I really like being in Edinburgh because of the city but also the quality of the people at the University. It’s very exciting to be among so many top scientists.” This article was published on 2023-11-10
