Tut MUVIS image retrieval system proposal for MSR-Bing challenge 2014

Author

Raitoharju, Jenni ; Zhang, Haijun ; Ozan, E.C. ; Waris, M.A. ; Faisal, Mohammad ; Cao, Guo-yun ; Roininen, M. ; Ahmad, Ishtiaq ; Shetty, R. ; C, S.P. ; Uhlmann, Stefan ; Samiee, Kaveh ; Kiranyaz, Serkan ; Gabbouj, Moncef

Author_Institution

Tampere Univ. of Technol., Tampere, Finland

fYear

2014

fDate

14-18 July 2014

Firstpage

1

Lastpage

6

Abstract

This paper presents our system designed for MSR-Bing Image Retrieval Challenge @ ICME 2014. The core of our system is formed by a text processing module combined with a module performing PCA-assisted perceptron regression with random sub-space selection (P²R²S²). P²R²S² uses Over-Feat features as a starting point and transforms them into more descriptive features via unsupervised training. The relevance score for each query-image pair is obtained by comparing the transformed features of the query image and the relevant training images. We also use a face bank, duplicate image detection, and optical character recognition to boost our evaluation accuracy. Our system achieves 0.5099 in terms of DCG₂₅ on the development set and 0.5116 on the test set.

Keywords

image retrieval; object detection; principal component analysis; regression analysis; MSR-Bing image retrieval challenge; P2R2S2; PCA-assisted perceptron regression; TUT MUVIS image retrieval system; descriptive features; duplicate image detection; face bank; optical character recognition; query-image pair; random subspace selection; text processing module; unsupervised training; Detectors; Face; Feature extraction; Optical character recognition software; Reliability; Training; Vectors; Data Partitioning; Face Bank; Image Retrieval; Relevance Evaluation;

fLanguage

English

Publisher

ieee

Conference_Titel

Multimedia and Expo Workshops (ICMEW), 2014 IEEE International Conference on

Conference_Location

Chengdu

ISSN

1945-7871

Type

conf

DOI

10.1109/ICMEW.2014.6890600

Filename

6890600