Vitæ
Since 2018
Reserach Intern, Microsoft, AI for Earth
2017
Visiting scholar at Shanghai Jiaotong University (上海交通大学), Shanghai,
China, at the group of Prof. Weiyao Lin.
2016
Visiting scholar at ICSI / UC Berkeley, California, USA, at the group of
Prof. Trevor Darrell.
2014 – 2018/19
PhD student in computer vision at Computer Vision Group,
Friedrich-Schiller-Universität, Jena, Germany.
Supervised by Prof. Dr. Joachim Denzler and Dr. Erik Rodner.
2012
Exchange Semester, University of Mississippi, Oxford, USA.
2011 – 2014
Master of science in computer science, Friedrich-Schiller-Universität, Jena,
Germany. Received “Germany Scholarship” for high-achieving and committed
students provided by IBM, demandware and the state of Germany.
2008 – 2011
Bachelor of science in computer science, Friedrich-Schiller-Universität, Jena,
Germany.
Research
DEEP LEARNING
Convolutional networks tremendously boosted the classification accuracy in the recent years. My goal is to exploit their inherent knowledge for other tasks like part discovery and localization. Our ACCV and ICCV paper published in 2014 and 2015 show the huge potential of this approach.
CLASSIFICATION
My work focuses on fine-grained classification, which is the automated recognition of visually very similar object categories like bird species or car models. I concentrate on improving state-of-the-art approaches especially in the scenario of weakly annotated images.
PART LEARNING
A core ingredient of most fine-grained classification approaches are object part locations. In order to improve classification accuracy, my research also includes finding the most discriminative object parts without the need of annotations provided by humans. This will allow for transferring the extensive knowledge of fine-grained recognition approaches to generic classification tasks.
APPLICATIONS
Besides pure research, I also contribute to the development of products at my chair’s partners in industry. A document classification system and a road segmentation system are examples for that.
Selected publications
Marcel Simon, Erik Rodner, Trevor Darrell, Joachim Denzler. The whole is more than its parts? From explicit to implicit pose normalization.
IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI). 2018.
Marcel Simon and Yang Gao and Trevor Darrell and Joachim Denzler and Erik Rodner. Generalized orderless pooling performs implicit salient matching.
International Conference on Computer Vision (ICCV). 2017.
Neural Activation Constellations: Unsupervised Part Model Discovery with Convolutional Networks
International Conference on Computer Vision (ICCV). 2015.
Fine-grained Recognition in the Noisy Wild: Sensitivity Analysis of Convolutional Neural Networks Approaches
British Machine Vision Conference (BMVC). 2016.
Part Detector Discovery in Deep Convolutional Neural Networks
Asian Conference on Computer Vision (ACCV). 162-177. 2014.
Fine-grained Classification of Identity Document Types with Only One Example
Machine Vision Applications (MVA). 126 – 129. 2015.
Fine-grained Recognition Datasets for Biodiversity Analysis
CVPR Workshop on Fine-grained Visual Classification (CVPR-WS). 2015
Convolutional Neural Networks as a Computational Model for the Underlying Processes of Aesthetics Perception
ECCV Workshop on Computer Vision for Art Analysis. 2016.
A list of all publications can be found here.