Julian Stöttinger


Senior Data Scientist at Verisk Analytics, former with SAP. Deep Learning, grammars and stuff.

Before that, I've been a post-doctoral researcher at the Knowdive group at the University of Trento, an associate center of the European Institute of Innovation and Technology. This is a private webpage to keep my research available.

Main research interests have been computer vision and machine learning. In particular, local features, color-based recognition, semantic image understanding, and recently, compositional approaches for visual recognition.

My Former Research Groups

KnowDive, University of Trento
Computer Vision Lab, Vienna University of Technology
Intelligent System Labs, University of Amsterdam
IMEDIA, INRIA Paris-Rocquencourt

Past FP7 Projects

CUbRIK - opens the blackbox of search
Glocal - event-based retrieval of networked media
Living Knowledge - diversity as an intrinsic and unavoidable property for knowledge
MUSCLE, Multimedia Understanding through Semantics, Computation and Learning


A Hybrid Machine-Crowd Approach to Photo Retrieval Result Diversification
Anca-Livia Radu, Bogdan Ionescu, Maria Menéndez, Julian Stöttinger, Fausto Giunchiglia, Antonella De Angeli, 20th International Conference on MultiMeida Modeling (MMM), Dublin, Ireland, 2014.

Ein Silikon Valley im Trentino - KURIER, 27.08.2012 (German) web

Social Events and Social Ties
Javier Paniagua, Ivan Tankoyeu, Julian Stöttinger, Fausto Giunchiglia, "Social Events in Web Multimedia" at the ACM International Conference on Multimedia Retrieval (ICMR), 2013.

Representativeness and Diversity in Photos via Crowd-Source Media Analysis
Anca-Livia Radu, Julian Stöttinger, Bogdan Ionescu, María Menendez, Fausto Giunchiglia, 10th International Workshop on Adaptive Multimedia Retrieval (AMR), 2012.

Key Publications

Indexing Media by Personal Events
Javier Paniagua, Ivan Tankoyeu, Julian Stöttinger, Fausto Giunchiglia, ICMR, 2012.
Dataset available on request.

(Unseen) Event Recognition via Composition
Julian Stöttinger, Jasper Uijlings, Anand K. Pandey, Nicu Sebe, Fausto Giunchiglia, CVPR, 2012.
Dataset available! 

Sparse Color Interest Points for Image Retrieval and Object Categorization
Julian Stöttinger, Allan Hanbury, Nicu Sebe, Theo Gevers, Transactions on Image Processing, 2012.
[download linux64 binary]

Color Based Skin Classification
Rehan Khan, Allan Hanbury, Julian Stöttinger, Abdul Bais, Pattern Recognition Letters, Volume 33, Issue 2, 2012, p 157 - 163.