MindGarage
| Motto | Open Minded Research |
|---|---|
| Established | October 2016 |
Field of research | Machine Learning, Deep Learning, Document analysis, Computer Vision |
| Address | Room 86-(101), TU Kaiserslautern |
| Location | Kaiserslautern, Rhineland Palatinate, Germany |
| 67663 | |
| Website | http://www.mindgarage.de/ |
The MindGarage[1] is an independent lab for research in the field of Machine Learning and Deep Learning, led by Marcus Liwicki. For the area of Kaiserslautern it is a unique prototype of research-industry collaboration where local industry supports modern teaching and research to strengthen the IT location without direct benefits in terms of software and projects. In particular, the MindGarage is mainly supported by the University of Kaiserslautern, where it is an official research group and by Insiders Technologies GmbH.[2]
The vision of the MindGarage is "Open Minded Research" and hence it is open to anyone for collaboration. The current research areas that MindGarage is involved in are Deep LearningGenerative Modelling, Image Recognition, Document Analysis, Natural Language Processing, Handwriting Recognition and Sentiment Analysis.
Teaching[edit]
The research lab is headed by its founder Marcus Liwicki and co-founder Muhammad Zeshan Afzal. Three lectures (Very Deep Learning, Applications of AI, and Natural Language Processing with Deep Learning) and a Seminar (MindStorm) form the teaching frame of the MindGarage. The lecture style is not in the typical frontal manner, but involves various discussions and in-depth analyses . In practical exercises, participants have access to the computing resources of the MindGarage.
Research[edit]
The laboratory focuses on both practical and theoretical aspects of Deep Neural Networks. It is part of its mission to produce open minded research in collaboration with other scientific fields. Specifically, current research production has focused on image generation using generative adversarial networks, as well as image classification and document analysis and recognition. References are given in the appendix.[3][4][5]
Collaborations[edit]
The research lab is currently hosted by the University of Kaiserslautern and works in close collaboration with Insiders Technologies GmbH and the University of Fribourg. The laboratory has additionally initiated collaborations with the Microeletronic Systems lab, the Machine Learning laband the Socioinformatics group at University of Kaiserslautern and the . Students of the University of Kaiserslautern that are required to take part in a research project are also entitled to do so in the MindGarage.
Team[edit]
The research team is mostly composed by students or graduates of the University of Kaiserslautern and of the University of Fribourg.
Prices & Awards[edit]
- The MindGarage won three tasks of the ICDAR2017 Competition on Multi-font and Multi-size Digitally Represented Arabic Text[6]
See also[edit]
References[edit]
- ↑ http://www.mindgarage.de
- ↑ "FB-Informatik: Arbeitsgruppen".
- ↑ * Afzal, Muhammad Zeshan, et al. "Cutting the Error by Half: Investigation of Very Deep CNN and Advanced Training Strategies for Document Image Classification." arXiv:1704.03557 (2017).
- ↑ Dash, Ayushman, et al. "AirScript-Creating Documents in Air." arXiv:1705.11181 (2017).
- ↑ Dash, Ayushman, et al. "TAC-GAN-Text Conditioned Auxiliary Classifier Generative Adversarial Network." arXiv:1703.06412 (2017).
- ↑ Slimane, Fouad; Ingold, Rolf; Hennebert, Jearn (2011-11-13). "ICDAR2017 Competition on Multi-font and Multi-size Digitally Represented Arabic Text". 2017 14th IAPR International Conference on Document Analysis and Recognition. 14.
This article "MindGarage" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:MindGarage. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
