CERTH at MediaEval Placing Task 2013

We describe the participation of the CERTH team in the Placing task of MediaEval 2013. We submitted 5 runs on the full test set, two of which are based on tag information, two on visual content, and one uses both tag and visual information. Our best performance (median error 650km) was achieved with the use of tag features.

[1]  Adam Rae,et al.  Working Notes for the Placing Task at MediaEval 2011 , 2011, MediaEval.

[2]  Steven Schockaert,et al.  Finding locations of flickr resources using language models and similarity search , 2011, ICMR.

[3]  Cordelia Schmid,et al.  Product Quantization for Nearest Neighbor Search , 2011, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Hervé Jégou,et al.  Negative Evidences and Co-occurences in Image Retrieval: The Benefit of PCA and Whitening , 2012, ECCV.

[5]  Yiannis Kompatsiaris,et al.  An empirical study on the combination of surf features with VLAD vectors for image search , 2012, 2012 13th International Workshop on Image Analysis for Multimedia Interactive Services.

[6]  Michael I. Jordan,et al.  Latent Dirichlet Allocation , 2001, J. Mach. Learn. Res..

[7]  Bart Thomee,et al.  Working Notes for the Placing Task at MediaEval 2013 , 2013, MediaEval.