Dr.-Ing. Peer Neubert
News and recent events
- In October'22 I became a professor at the University in Koblenz.
- Our paper SEER: Unsupervised and sample-efficient environment specialization of image descriptors has been accepted for RSS'22.
- Our paper HDC-MiniROCKET: Explicit Time Encoding in Time Series Classification with Hyperdimensional Computing has been accepted for IJCNN'22.
- Our paper A comparison of Vector Symbolic Architectures has been accepted for Springer Artificial Intelligence Review journal. Congratulations to the first author Kenny Schlegel!
- I've been nominated for the Heinz Maier-Leibnitz Preis 2022 (Deutsche Forschungsgemeinschaft).
- Both submissions for RSS'21 have been accepted. My own paper "Vector Semantic Representations as Descriptors for Visual Place Recognition" and Stefan Schubert's paper (co-authored by me) "Fast and Memory Efficient Graph Optimization via ICM for Visual Place Recognition".
- I'm co-author of a paper by Kenny Schlegel and Florian Mirus (BMW) on Multivariate Time Series Analysis for Driving Style Classification using Neural Networks and Hyperdimensional Computing that was accepted at IEEE IV'21.
- The paper Hyperdimensional computing as a framework for systematic aggregation of image descriptors co-oauthored with Stefan Schubert was accepted for CVPR'21.
- All three papers that I co-authored have been accepted for ICRA'21: Stefan Schubert's paper "Beyond ANN: Exploiting Structural Knowledge for Efficient Place Recognition", Fangming Yuan's paper "SoftMP: Attentive feature pooling for joint local feature detection and description for place recognition in changing environments", and my own combined RA-L/ICRA paper below. Congratulation and many thanks to Stefan and Fangming!
- The paper "Resolving Place Recognition Inconsistencies Using Intra-Set Similarities" together with Stefan Schubert and Peter Protzel will appear in IEEE RA-L'21.
- I'm co-author of Stefan Schubert's paper "Graph-based Non-Linear Least Squares Optimization for Visual Place Recognition in Changing Environments" that will appear in IEEE RA-L'21.
We organized a free online tutorial on Vector Symbolic Architectures at the European Conference on Artificial Intelligence 2020 (ECAI). More information, slides, and code can be found on our tutorial website.
Research Interests
My research interests are in the fields of autonomus robotis asn systems as well as artificial intelligence. I use computer vision and machine learning as tools within these fields. I have particular experience in the areas of place recognition in changing environments, hand crafted and deep-learned visual features, mobile robot navigation, biologically inspired perception and navigation approaches, as well as hyperdimensional computing. Details on my research can be found in the list of publications or here:- Vector Symbolic Architectures
- Place Recognition in Changing Environments
- Superpixel Segmentations
- Camera-based Navigation in (semantic) 3D Maps
- Circular Convolutional Neural Networks (CCNNs)
- How to Build and Customize a High-Resolution 3D Laserscanner Using Off-the-shelf Components
- Robotic Systems
Publications
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()