Matthias Springstein (M.Eng.)

Research Assistant

Email: matthias.springsteintibeu

Phone: +49 511 762-19916

Postal address: Welfengarten 1 B, 30167 Hannover

Address for visitors: Lange Laube 28, 30159 Hannover, Room: 2.09

Matthias Springstein

My research focuses on web-supervised learning regarding visual concepts and incremental learning approaches. The objective is to minimize efforts for manually labeling training data which might also enable and improve domain adaptation for visual concept models. This requires, for instance, automatic acquisition of visual information from user-generated content such as Flickr and YouTube. In general, I am interested in deep learning, concept detection, and training data augmentation.

  • M. Mühling, N. Korfhage, E. Müller, C. Otto, M. Springstein, T. Langelage, U. Veith, R. Ewerth, B. Freisleben: 
    Deep learning for content-based video retrieval in film and television production. 
    In Multimedia Tools and Applications, 2017, 1-26. 
    https://link.springer.com/article/10.1007/s11042-017-4962-9

  • R. Ewerth, M. Springstein, E. Müller, A. Balz, J. Gehlhaar, T. Naziyok, K. Dembczyński, E. Hüllermeier:
    Estimating Relative Depth in Single Images via Rankboost
    In: Proceedings of IEEE 18th International Conference on Multimedia & Expo (ICME), Hongkong, IEEE Press, 2017, 919-924.
    https://doi.org/10.1109/ICME.2017.8019434

  • E. Müller, M. Springstein, R. Ewerth: 
    "When was this picture taken?" – Image Date Estimation in the Wild
    In: Proceedings of 39th European Conference on Information Retrieval (ECIR), Aberdeen, UK, 2017, 619-625.
    https://link.springer.com/chapter/10.1007/978-3-319-56608-5_57

  • R. Ewerth, M. Springstein, L.A. Phan-Vogtmann, J. Schütze: 
    "Are Machines Better in Image Tagging?" – A User Study Adds to the Puzzle
    In: Proceedings of 39th European Conference on Information Retrieval (ECIR), Aberdeen, UK, 2017, 186-198.
    https://link.springer.com/chapter/10.1007/978-3-319-56608-5_15

  • M. Springstein, R. Ewerth: 
    On the Effects of Spam Filtering and Incremental Learning for Web-supervised Visual Concept Classification
    In: Proceedings of ACM International Conference on Multimedia Retrieval (ICMR), New York, ACM, 2016, 377-380
    http://dl.acm.org/citation.cfm?id=2912072