Guillermo Gallego has been Professor of “Robotic Interactive Perception” at Technische Universität Berlin and Einstein Center Digital Future (ECDF) since September 2019. The Spaniard previously worked as a postdoctoral fellow at the Department of Neuroinformatics of the University of Zurich and ETH Zurich. From 2005 to 2011, he was a Fulbrighter conducting research at the Georgia Institute of Technology in the USA, where he received his doctorate in 2011.
The focus of his research work lies on at the intersection of robotics, computer vision and applied mathematics, with an emphasis on optimization. “Broadly speaking, my focus is on robot perception using cameras, and on optimization methods for interdisciplinary imaging and control problems”, explains Gallego.
Within the framework of his ECDF professorship, he wants to focus on improving the perception system in robots. “As sensors (such as cameras) and processors become cheaper, smaller and better, they also become more ubiquitous. This is something that we observe in everyday life, on our smartphones. It is also the case of robots and remote sensing networks”, says Gallego.
The amount of information that such devices are able to acquire about the environment is enormous and steadily growing. To make the most out of these technological advantages, new efficient methods are needed to extract information (patterns) from the raw data and convert it into knowledge. “Such perceptual knowledge is essential to make predictions about the future or to provide autonomy for robots operating in changing environments”, says Gallego. This is a challenging research project that involves multiple trade-offs, since the amount of available resources like memory, computing capabilities, power and bandwidth is largely constrained.
Gallego is very interested in interdisciplinary research. As an example, he cites a research project during his doctoral studies. “I combined my electrotechnical background with that from researchers in civil engineering to design a remote sensing tool that allowed us to extract information from stereo videos of the surface of the ocean. We used it to study nature, such as to measure the energy-transfer interaction between the sea and the atmosphere over large areas and in a non-intrusive way, and to provide an early warning system for coastal regions at risk, such as the Venice Lagoon, Italy. I am sure that similar interesting opportunities will arise among the ECDF professors due to our diversified knowledge.”
Other interdisciplinary fields of research are digital signal processing and machine learning. “These fields are boundless, with applications in many scientific disciplines. Everyday, huge amounts of data, in the form of images or videos, are generated, and almost anywhere where there is an image it is digital and calls to be processed to extract meaningful information from it. We, as humans, do it effortlessly, but for computers it is another story; mathematical techniques from the above fields, now under the name of Data Science, are used to extract such information in a computer. Having mathematics and information processing at my heart, it opens many possibilities to apply this knowledge to other disciplines, as far apart as medicine, autonomous driving, space exploration or social media.“