AXES @ TRECVID 2014: Instance Search

The AXES project submitted three runs for instance search at TRECVid 2014 [1]: two interactive runs and one automatic run. These runs used exactly the same instance search engine as was used by AXES in 2013, but incorporate a custom interface for the instance search tasks, pseudo relevance feedback, and result list expansion. Unfortunately, these changes did not improve accuracy significantly over our 2013 results in the interactive task.

[1]  Andrew Zisserman,et al.  On-the-fly specific person retrieval , 2012, 2012 13th International Workshop on Image Analysis for Multimedia Interactive Services.

[2]  Paul Over,et al.  Evaluation campaigns and TRECVid , 2006, MIR '06.

[3]  C. Schmid,et al.  On the burstiness of visual elements , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[4]  Cordelia Schmid,et al.  Improving Bag-of-Features for Large Scale Image Search , 2010, International Journal of Computer Vision.

[5]  Cordelia Schmid,et al.  The AXES PRO video search system , 2013, ICMR '13.

[6]  Georges Quénot,et al.  TRECVID 2015 - An Overview of the Goals, Tasks, Data, Evaluation Mechanisms and Metrics , 2011, TRECVID.

[7]  Cordelia Schmid,et al.  The AXES research video search system , 2014, ICASSP 2014.

[8]  Andrew Zisserman,et al.  VISOR: Towards On-the-Fly Large-Scale Object Category Retrieval , 2012, ACCV.

[9]  Cordelia Schmid,et al.  AXES at TRECVID 2012: KIS, INS, and MED , 2012, TRECVID.