This hands-on course will take you from 0 to 100 in Deep Learning with Keras. Our aim is to teach the fundamentals of deep learning with Convolutional Neural Networks (CNN) based on modern techniques using the Keras API and the Tensorflow backend. By the end participants will know how to build deep learning models, how to train them, what to avoid during training, what to check during training and how to perform model inference, especially for image based problems. We hope participants will then go out and apply these methods to their own problems and use cases.
The core curriculum is planned from Monday (September 24) to Friday afternoon (September 28) to take place at the MPI CBG campus, Pfotenhauerstrasse 108, Dresden, Germany. As the agenda is currently being prepared please check-in from time to time.
All participants are expected to bring their laptop. During the workshop, a uniform access to GPU-enabled workstations or servers will be provided that hold the software stack used. Thus, your laptop is not required to hold a mobile GPU or alike. All participants are expected to have a solid understanding of fundamentals of linear algebra as well as programming.
The workshop admission fee amounts to € 250 per participant. Every successfull applicant is required to bring a poster to the workshop that describes their current scientific challenge that they would like to solve with Deep Learning. Posters have to be sent in 1 week prior to the workshop.
This year, we are happy to have the following individuals as instructors:
- Walter de Back (TU Dresden, Universitätsklinikum Dresden)
Walter de Back has a MSc in Artificial Intelligence from Utrecht University (NL) and worked as a Junior Fellow at the the Collegium Budapest-Institute for Advanced Study. He studied pattern formation in tissues using multi-scale agent-based simulations, for which he obtained a PhD in Computational Biology from TU Dresden. Currently, Walter works as a postdoc data scientist at the Faculty of Medicine (TU Dresden) where he uses deep learning for biomedical image analysis. Recent projects include cell segmentation in live cell microscopy, dental age estimation from panoramic radiographs, and tumor tissue classification based on mass spectrometry imaging data.
Jeffrey Kelling (HZDR)
Jeffrey Kelling obtained his diploma in statistical physics and his Ph.D. in physics on massively parallel lattice Monte-Carlo simulations on GPUs. He is a scientist in the computational science group at the Helmholtz-Zentrum Dresden-Rossendorf, concerned with high performance computing and deep learning applications in science. One of his current topics is using real-time object detection to detect problems in pulsed high-power lasers.
Thomas Neumann (Freelancer)
Thomas Neuman is a free-lance R&D software developer. Until August 2018, he was a PostDoc at the Junior Research Group "TISRA" at Hochschule für Technik und Wirtschaft (HTW) Dresden. His research interests range from 3D reconstruction of dynamic 3D surfaces in particular to machine learning for computer vision as well as statistical methods to model 3D surfaces. He defended his PhD in 2016 at the Institute for Computer Graphics at TU Braunschweig on "Reconstruction,
Analysis, and Editing of dynamically deforming 3D-Surfaces".
- Kashif Rasul (Zalando Research)
- Uwe Schmidt (MPI CBG)
Uwe Schmidt received the MSc and PhD degrees in computer science from TU Darmstadt, Germany. He has been a visiting graduate student at the University of British Columbia in Vancouver, Canada. Uwe is currently a postdoc at MPI-CBG in Dresden, Germany. His research interests include machine learning and computer vision.
- Steffen Seitz (TU Dresden)
Steffen studied Electrical Engineering and Nanobiophysics receiving his Dipl. Ing./M.Sc. from TU Dresden in 2016 in the field of Information Theory.
Since 2016 he is working towards his Ph.D. using RNN-Autoencoders as a novel failure forecasting approach in Industrial Prognostics and Health Management Systems at TU-Dresden.
The key idea is to train Autoencoders to detect good samples and using their reconstruction error on the degenerated samples as an universal feature to forecast the system Health. Therefore he successfully implemented various Autoencoder architectures in the past using Keras and Tensorflow. Currently he is focussing to implement STORN Autoencoders and DVBF-Kalman filters on the task.
Sebastian Starke (HZDR)
Sebastian received his bachelor degree in mathematics in 2013 and his masters degree in statistics in 2015 from the Otto-von-Guericke University in Magdeburg. Afterwards he worked as an algorithm engineer in the field of speech recognition before joining the computational science group at HZDR in October 2016. At the moment he is working together with OncoRay scientists to apply deep learning methods to CT images of cancer patients to improve personalized treatment.
Martin Weigert (MPI CBG)
Martin Weigert holds a Diploma in Physics from Technical University Dresden. He is currently wrapping up his PhD in the group of Gene Myers at MPI-CBG in Dresden, where he investigates computational methods for advanced fluorescence microscopy. Among his interests are computational optics, physical simulations and visualizations, and machine learning methods for image reconstruction.