Tomas Gerlich is a PhD Candidate in Computer Science at the University of Illinois at Chicago where he conducts research in computer vision and mobile sensing.
I am interested in various aspects of computer vision. My dissertation research concerns the design of novel optical flow algorithms to automatically interpret motion in traffic scene videos. This research has many applications in transportation, for instance in vehicle counting, speed estimation, signal optimization, traffic management and surveillance. I approach the optical flow problem from the (projective) geometry viewpoint and often leverage structure-from-motion (SfM) parametrizations in reasoning about rigid multi-motion.
My research is also in the domain of mobile sensing. The widespread emergence of GPS and accelerometer sensors embedded within smartphones lead us to develop many novel applications for transit tracking and route planning. For instance, the collection of long-term GPS traces which correspond to bus trajectories enabled us to develop algorithms capable of estimating a complete transit system along with route shapes, bus stop locations, a daily schedule and any subsequent real-time map-matching and bus tracking. This low-cost and easy-to-deploy system is especially suitable for small transit agencies similar in scale to a university campus-wide bus sytem.
I am also interested in deep-learning, machine learning and artificial intelligence in general. I enjoy the field of computational photography and in the future I would like to tackle many of the computer vision problems early at the image capture level. I would like to design sensors, external illumination and algorithms and explore their advantages in simplifying any given computer vision task, such as computing optical flow.
Tomas Gerlich and Jakob Eriksson.
Optical Flow for Rigid Multi-Motion Scenes.
In Fourth International Conference on 3D Vision (3DV), pp. 212-220, Stanford CA, 2016.
James Biagioni, Tomas Gerlich, Timothy Merrifield, and Jakob Eriksson.
EasyTracker: Automatic Transit Tracking, Mapping, and Arrival Time Prediction Using Smartphones.
In Embedded Networked Sensor Systems (SenSys), pp. 68-81, Seattle WA, 2011.
Tomas Gerlich, James Biagioni, Timothy Merrifield, and Jakob Eriksson.
Demo: Tracking Transit with EasyTracker.
In Embedded Networked Sensor Systems (SenSys), pp. 401-402, Seattle WA, 2011.
Arvind Thiagarajan, James Biagioni, Tomas Gerlich, and Jakob Eriksson.
Cooperative Transit Tracking using GPS-enabled Smartphones.
In Embedded Networked Sensor Systems (SenSys), pp. 85-98, Zurich, Switzerland, 2010.
James Biagioni, Adrian Agresta, Tomas Gerlich, and Jakob Eriksson.
TransitGenie: A real-time, context-aware transit navigator (demo abstract).
In Embedded Networked Sensor Systems (SenSys), pp. 329-330, Berkeley CA, 2009.
I earned my Bachelor of Science in Computer Science from the University of Illinois at Chicago in 2009. I began my Ph.D. studies in the same department in the Fall of 2009. I work in the BITS Networked Systems Laboratory under the guidance of Prof. Jakob Eriksson.