Dipl.-Math. techn. Florian Kuhnt
Wissenschaftlicher Mitarbeiter
Werdegang
Florian Kuhnt studierte bis 2012 Technomathematik am Karlsruher Institut für Technologie (KIT). Seine Schwerpunkte liegen in den Gebieten Kognitive Systeme und Robotik.
Seine Diplomarbeit mit dem Titel „Probabilistische Kollisionsprädiktion für Segway-Transporter“ führte er in der Abteilung Technisch Kognitive Systeme (TKS) am Forschungszentrum Informatik (FZI) durch.
Seit Juni 2012 ist er wissenschaftlicher Mitarbeiter in der Abteilung TKS. Er ist im Bereich der Lokalisierung und Prädiktion von Verkehrsteilnehmern unter Berücksichtigung von Karteninformationen tätig.
Seit 2016 ist er Projektleiter für verschiedene öffentlich geförderte und industriell beauftragte Projekte, z.B.:
- EVA Shuttle Busse
- RobustSENSE
- tech center a-drive
- Audi Autonomous Driving Cup
Publikationen
Eine Liste aller Publikationen gibt es auf researchgate.
Konferenzbeitrag (10)
- Integrating End-to-End Learned Steering into Probabilistic Autonomous DrivingInfoDetails
Christian Hubschneider and André Bauer and Jens Doll and Michael Weber and Sebastian Klemm and Florian Kuhnt and J. Marius Zöllner, 2017
In this paper, we propose an integrated approach of combining end-to-end learned trajectory proposals with a probabilistic sampling based planning algorithm for autonomous driving. A convolutional neural network is trained based on monocular image data to predict prospective steering angles. By using a local history of image data, we achieve an implicit spatial representation of parking cars or other obstacles commonly found in urban and residential areas. Through this local history, calculated using the vehicle's velocity data, the trajectory proposals are not only capable of lane following, but also comfortably circumnavigate obstacles. Training data is collected by recording video data and the vehicles CAN bus during human driving, thus imitating human behavior. The integration of end-to-end learning into a modularized architecture allows for additional safety constraints and complementary sensor information to be combined with intuitive steering. Our first results take a promising step towards general architectures for autonomous vehicles that combine deep learning with factorized probabilistic modeling.
- Automated Vehicle System Architecture with Performance AssessmentDetails
Tas, \"Omer Sahin and Hörmann, Stefan and Schäufele, Bernd and Kuhnt, Florian, 2017
- Integrating End-to-End Learned Steering into Probabilistic Autonomous DrivingInfoDetails
Christian Hubschneider and André Bauer and Jens Doll and Michael Weber and Sebastian Klemm and Florian Kuhnt and J. Marius Zöllner, 2017
In this paper, we propose an integrated approach of combining end-to-end learned trajectory proposals with a probabilistic sampling based planning algorithm for autonomous driving. A convolutional neural network is trained based on monocular image data to predict prospective steering angles. By using a local history of image data, we achieve an implicit spatial representation of parking cars or other obstacles commonly found in urban and residential areas. Through this local history, calculated using the vehicle's velocity data, the trajectory proposals are not only capable of lane following, but also comfortably circumnavigate obstacles. Training data is collected by recording video data and the vehicles CAN bus during human driving, thus imitating human behavior. The integration of end-to-end learning into a modularized architecture allows for additional safety constraints and complementary sensor information to be combined with intuitive steering. Our first results take a promising step towards general architectures for autonomous vehicles that combine deep learning with factorized probabilistic modeling.
- Understanding Interactions between Traffic Participants based on Learned BehaviorsDetails
Kuhnt, Florian and Schulz, Jens and Schamm, Thomas and Zöllner, J Marius, 2016
- Functional System Architectures towards Fully Automated DrivingDetails
Tas, \"Omer Sahin and Kuhnt, Florian and Zöllner, J. Marius and Stiller, Christoph, 2016
- Testing and Validating High Level Components for Automated Driving : Simulation Framework for Traffic ScenariosDetails
Zofka, Marc Ren\'e and Klemm, Sebastian and Kuhnt, Florian and Schamm, Thomas and Zöllner, J. Marius, 2016
- DSRC and Radar Object Matching for Cooperative Driver Assistance SystemsDetails
Chen, Qi and Yuan, Ting and Gern, Axel and Roth, Tobias and Kuhnt, Florian and Breu, Jakob and Bogdanovic, Miro and Weiss, Christian, 2015
- Towards a Unified Traffic Situation Estimation Model – Street-dependent Behaviour and Motion Models –Details
Kuhnt, Florian and Kohlhaas, Ralf and Schamm, Thomas and Zöllner, J. Marius, 2015
- Data-Driven Simulation and Parametrization of Traffic Scenarios for the Development of Advanced Driver Assistance SystemsDetails
Zofka, Marc Ren\'e and Kuhnt, Florian and Kohlhaas, Ralf and Rist, Christoph and Schamm, Thomas and Zöllner, J. Marius, 2015
- Particle filter map matching and trajectory prediction using a spline based intersection modelDetails
Kuhnt, Florian and Kohlhaas, Ralf and Jordan, Rüdiger and Gu\ssner, Thomas and Gumpp, Thomas and Schamm, Thomas and Zollner, J. Marius, 2014
Thesis (1)
- Probabilistische Kollisionsprädiktion für Segway-TransporterDetails
Kuhnt, Florian, 2011
Betreute Thesis (4)
- Development of Data Matching Algorithms for Vehicle Tracking by Car2Car Communication and Radar SystemsDetails
Roth, Tobias, 2015
- Erkennung von Interaktionen zwischen Verkehrsteilnehmern zur VerhaltensprädiktionDetails
Schulz, Jens, 2015
- Fusion von Laserscanner- und Tiefenbildkameradaten zur Umweltmodellierung für den Versuchsträger CoCarDetails
Benno Evers, 2014
- Situationsabhängige Prädiktion von Fahrzeugbewegungen in urbaner Umgebung für FahrerassistenzsystemeDetails
Patrick Bartler, 2014
Export Suchergebnis .bib
Kontakt
Telefon: +49 721 9654-364
E-Mail: kuhnt@ fzi.de