Dr Tobias Fischer conducts interdisciplinary research at the intersection of intelligent robotics, computer vision, and computational cognition. His main goal is to develop high-performing, bio-inspired computer vision algorithms that simultaneously examine animals/humans and robots' perceptional capabilities. Before joining the Queensland University of Technology's Centre for Robotics as an Associate Investigator and Research Fellow in January 2020, Dr Fischer was a postdoctoral researcher in the Personal Robotics Lab at Imperial College London. He received a PhD from Imperial College in January 2019. Tobias' thesis was awarded the UK Best Thesis in Robotics Award 2018 and the Eryl Cadwaladr Davies Award for the best thesis in Imperial's EEE Department in 2017-2018. He previously received an M.Sc. degree (distinction) in Artificial Intelligence from The University of Edinburgh in 2014 and a B.Sc. degree in Computer Engineering from Ilmenau University of Technology, Germany, in 2013. His works have attracted two best poster awards, one best paper award, and he is the senior author of the winning submission to the Facebook Mapillary Place Recognition Challenge 2020.

Nature has long been an inspiration for robotic systems, both in terms of their physical design as well as in terms of the algorithms that robots make use of. In this talk, Tobias will provide an overview of recent advances in neuromorphic computing, an emerging area within artificial intelligence that emulates the structure of the human brain. Neuromorphic computing promises sensors, algorithms and processors that are much more energy efficient compared to their conventional counterparts, while being able to adapt rapidly to new incoming information without forgetting previous knowledge, which is an issue for a range of other algorithms. Tobias will focus on two pieces of his research: The first piece of research makes use of bio-inspired event cameras that capture changes in the environment on a pixel-by-pixel basis, without ever capturing all pixels at once. These cameras have properties that are beneficial in challenging environments: there is basically no motion blur, and event cameras have a very high dynamic range. As there is no concept of an image anymore, Tobias proposes to combine the events over a multitude of different time scales. The second piece of research makes use of spiking neural networks that closely mimic the neurons found in the brain. We train the network to distinguish between images that depict different places, even when the appearance of the place has changed significantly because of e.g. different seasons. This research is a significant milestone towards spiking neural networks that can provide robust, energy-efficient, and low latency robot localization.