Alexander Gepperth
Présentation
Personal Information Family status: married, 2children Nationality: German Experience in academia 5/2011 – present Tenured professor at „École Nationale Supérieure de Techniques Assignments: Teaching, research, supervision of theses, Research focus: large-scale learning in intelligent vehicles Industrial experience 02/2008 – 5/2011 Premature tenure as „Senior Scientist“ at Honda Research 10/2006 – 2/2008 Four-year contract as „Senior Scientist“ at Honda Research Assignments: Basic research in machine learning for intelligent PhD thesis 11/2002 – 04/2006 at the university of Bochum, institute for neural computation Subject: „Neural learning methods for visual object recognition“ Degree: Dr. rer. nat (grade: „very good“) Assignments: research, teaching, participation in third-party funded projects (Honda Research Institute Europe GmbH, Robert Bosch KG , DFG Sonderforschungsbereich 475) Tertiary education 10/1996 – 01/2002 Studies in physics at Ludwig-Maximilians-Universität Munich Diploma thesis: „Non-BPS states in string theory” (grade: 1,7) Degree: diploma (grade: „very good“) Alternate civil service (instead of army service) 8/1995 - 10/1996 at the municipal hospital Pfaffenhofen/Ilm Secondary education 6/1995 at Schyren-Gymnasium Pfaffenhofen/Ilm, grade: 1,8 Skills Computers Programming: C/C++, CUDA, Python, Matlab Web programming: HTML, CSS, PHP Operating systems: Windows, Linux Real-time middleware: ROS Scientific standard tools: LaTeX, svn, git, doxygen, bash, eclipse, gnuplot, make, cmake, ... Libraries: OpenCV, Qt, numpy/scipy, matplotlib/pylab Languages German, Czech: mother tongues English, French: fluent Spanish: advanced level Japanese:basic level Interests Tennis, volleyball, bodybuilding, Go, playing the violin, real-time strategy games (Starcraft)
Avancées“ (Palaiseau, France)
organization of the specialization subject „intelligent vehicles“
Institute Europe GmbH
Institute Europe GmbH, Offenbach am Main, Germany
vehicles, implementation of prototypes and demonstrations,
communication to management and academic community,
supervision of students
Publications
Publications
|
|
Marginal Replay vs Conditional Replay for Continual LearningICANN, 2019, Munich, Germany |
|
|
Towards incremental deep learning: multi-level change detection in a hierarchical recognition architectureEuropean Symposium on Artificial Neural Networks (ESANN), 2016, Bruges, Belgium |
|
|
Incremental Learning for Bootstrapping Object Classifier ModelsIEEE International Conference On Intelligent Transportation Systems (ITSC), 2016, Seoul, South Korea |
|
|
Learning to be attractive: probabilistic computation with dynamic attractor networksInternal Conference on Development and LEarning (ICDL), 2016, Cergy-Pontoise, France |
|
|
Incremental learning algorithms and applicationsEuropean Symposium on Artificial Neural Networks (ESANN), 2016, Bruges, Belgium |
|
|
Computational Advantages of Deep Prototype-Based LearningInternational Conference on Artificial Neural Networks (ICANN), 2016, Barcelona, Spain. pp.121 - 127, ⟨10.1007/978-3-319-44781-0_15⟩ |
|
|
A Deep Learning Approach for Hand Posture Recognition from Depth DataInternational Conference on Artificial Neural Networks (ICANN), 2016, Barcelona, Spain. pp.179 - 186, ⟨10.1007/978-3-319-44781-0_22⟩ |
|
|
A simple technique for improving multi-class classification with neural networksEuropean Symposium on artificial neural networks (ESANN), Jun 2015, Bruges, Belgium |
|
|
A light-weight real-time applicable hand gesture recognition system for automotive applicationsIEEE International Symposium on Intelligent Vehicles (IV), Jun 2015, Seoul, South Korea. pp.336-342, ⟨10.1109/IVS.2015.7225708⟩ |
|
|
Active learning of local predictable representations with artificial curiosityInternational Conference on Development and Learning and Epigenetic Robotics (ICDL-Epirob), Aug 2015, Providence, United States |
|
|
Resource-efficient incremental learning in very high dimensionsEuropean Symposium on Artificial Neural Networks (ESANN), Apr 2015, Bruges, Belgium |
|
|
Learning of local predictable representations in partially learnable environmentsThe International Joint Conference on Neural Networks (IJCNN), Jul 2015, Killarney, Ireland |
|
|
Using self-organizing maps for regression: the importance of the output functionEuropean Symposium on Artificial Neural Networks (ESANN), Apr 2015, Bruges, Belgium |
|
|
A generative-discriminative learning model for noisy information fusionInternational Conference on Development and Learning (ICDL), Aug 2015, Providence, United States. ⟨10.1109/DEVLRN.2015.7346148⟩ |
|
|
Biologically inspired incremental learning for high-dimensional spacesJoint IEEE International Conference Developmental Learning and Epigenetic Robotics (ICDL-EPIROB), Sep 2015, Providence, United States. ⟨10.1109/DEVLRN.2015.7346155⟩ |
|
|
Scene context is more than a Bayesian prior: Competitive vehicle detection with restricted detectorsIEEE International Symposium on Intelligent Vehicles(IV), May 2014, Detroit, United States. pp.1358 - 1364, ⟨10.1109/IVS.2014.6856542⟩ |
|
|
Latency-Based Probabilistic Information Processing in Recurrent Neural HierarchiesInternational Conference on Artificial Neural Networks (ICANN), Sep 2014, Hamburg, Germany. pp.715 - 722, ⟨10.1007/978-3-319-11179-7_90⟩ |
|
|
PROPRE: PROjection and PREdiction for multimodal correlations learning. An application to pedestrians visual data discriminationIJCNN - International Joint Conference on Neural Networks, Jul 2014, Pékin, China |
|
|
Context-based vector fields for multi-object tracking in application to road trafficIEEE International Conference On Intelligent Transportation Systems (ITSC), Oct 2014, Qingdao, China. pp.1179 - 1185, ⟨10.1109/ITSC.2014.6957847⟩ |
|
|
A real-time applicable 3D gesture recognition system for automobile HMIIEEE International Conference On Intelligent Transportation Systems (ITSC), Oct 2014, Qingdao, China. pp.2616 - 2622, ⟨10.1109/ITSC.2014.6958109⟩ |
|
|
Neural Network Based Data Fusion for Hand Pose Recognition with Multiple ToF SensorsInternational Conference on Artificial Neural Networks (ICANN), Sep 2014, Hamburg, Germany. pp.233 - 240, ⟨10.1007/978-3-319-11179-7_30⟩ |
|
|
Neural network based 2D/3D fusion for robotic object recognitionEuropean Symposium on artificial neural networks (ESANN), May 2014, Bruges, Belgium. pp.127 - 132 |
|
|
Robust visual pedestrian detection by tight coupling to trackingIEEE International Conference On Intelligent Transportation Systems (ITSC), Oct 2014, Qingdao, China. pp.1935 - 1940, ⟨10.1109/ITSC.2014.6957989⟩ |
|
|
Time-of-Flight based multi-sensor fusion strategies for hand gesture recognitionIEEE International Symposium on Computational Intelligence and Informatics, Nov 2014, Budapest, Hungary |
|
|
Latency-based probabilistic information processing in a learning feedback hierarchyInternational Joint Conference on Neural Networks (IJCNN), Jun 2014, Beijing, China. pp.3031 - 3037, ⟨10.1109/IJCNN.2014.6889919⟩ |
|
|
Discrimination of visual pedestrians data by combining projection and prediction learningEuropean Symposium on artificial neural networks (ESANN), Apr 2014, Bruges, Belgium |
|
|
A Multi-Modal System for Road Detection and SegmentationIEEE Intelligent Vehicles Symposium, Jun 2014, Dearborn, Michigan, United States. pp.1365-1370 |
|
|
Multimodal space representation driven by self-evaluation of predictabilityJoint IEEE International Conference Developmental Learning and Epigenetic Robotics (ICDL-EPIROB), Oct 2014, Gênes, Italy |
|
|
Neural Network Fusion of Color, Depth and Location for Object Instance Recognition on a Mobile RobotSecond Workshop on Assistive Computer Vision and Robotics (ACVR), in conjunction with European Conference on Computer Vision, Sep 2014, Zurich, Switzerland |
|
|
A Comparison of Geometric and Energy-Based Point Cloud Semantic Segmentation Methods6th European Conference on Mobile Robotics (ECMR), IEEE, Sep 2013, Barcelona, Spain. pp.88-93, ⟨10.1109/ECMR.2013.6698825⟩ |
|
|
Simultaneous concept formation driven by predictabilityInternational conference on development and learning, Nov 2012, San Diego, United States |
|
|
Co-training of context models for real-time object detectionIEEE Symposium on Intelligent Vehicles, Jun 2012, Madrid, Spain |
|
|
RGBD object recognition and visual texture classification for indoor semantic mappingTechnologies for Practical Robot Applications (TePRA), 2012 IEEE International Conference on, Apr 2012, United States. pp.127 - 132, ⟨10.1109/TePRA.2012.6215666⟩ |
|
|
New learning paradigms for real-world environment perceptionMachine Learning [cs.LG]. Université Pierre & Marie Curie, 2016 |