Challenge-based human-inspired computer vision

Join our workshops at FAU - Erlangen, BME - Budapest, and SSSA - Pisa to learn about human-inspired computer vision. Experience international and interdisciplinar collaboration, and participate in challenges to see whose solution performs better.

• The FAU workshop will deal with tracking hexbugs, miniature robots that move voluntarily on a field. We will discover basic ways for computer vision and will deal with contemporary convolutional neural networks. In an intense hands-on session, you will compile your own tracking algorithm. We record together a new test set that your algorithm hasn’t seen before and determine whose algorithm performs best in recognizing the hexbugs.
• The workshop at BME will aim at motion tracking based on different camera types in terms of spectral sensitivity, including non-visual spectral ranges. The goal is to compare various spectral domains for the tasks in order to optimize recognition and tracking performance. The expected outcome is a better understanding of the effect of spectral sensitivity on image quality and widened knowledge to enable future engineers to select a problem-specific optimal device.
• The workshop at SSSA will aim at experimenting mechatronic platforms that integrate vision, ultrasound and tactile sensors for characterizing the biomechanical properties of biological specimens, as well as to provide haptic feedback and multisensory telepresence to remote users. The expected outcome is to gain competences on how to integrate multisensory data, and to apply algorithms and AI strategies in the processing of information gathered by such platforms.