gaips_bea image1 image2 image3 image4 image5 gaips_ecute_beach_bar_banner gaips_ecute_train_incorrect_ticket_banner
HEIDI (HEIGH Dimensional Indexing) or how to index one billion of vectors Webpage icon

Collaborators: Andreas Wichert, Angelo Cardoso, Joao Sacramento, Catarina Pinto Moreira
Keywords: High dimensioal indexing, Curse of dimensionality


Traditional indexing of multimedia data leads to dilemma. Either the number of features has to be reduced or the quality of the results in unsatisfactory, or approximate queries is preformed leading to a relative error during retrieval. The promise of the recently introduced subspace-tree [1-5] is the logarithmic retrieval complexity of extremely high dimensional features. The subspace-tree indicates that the conjecture "the curse of dimensionality" is false. The search in such a structure starts at the subspace with the lowest dimension. In this subspace, the set of all possible similar objects is determined. In the next subspace, additional metric information corresponding to a higher dimension is used to reduce this set. This process is then repeated. The theoretical estimation of temporal complexity of the subspace tree is logarithmic for evenly distributed data. It means that depending on the distribution of our data, we have to choose an ideal projection into the subspaces leading to an ideal hierarchy. In the project we will perform experiments on different databases, high dimensional data up to several thousand, and of size till 1 billion.


Projection Based Operators in lp space for Exact Similarity Search
pdf iconbib icon
Andreas Wichert and Catarina Moreira, Fundamenta Informaticae: Annales Societatis Mathematicae Polonae, Vol. 136, No. 4, pg. 461-474, 2015
Finding Academic Experts on a MultiSensor Approach using Shannon's Entropy
pdf iconbib icon
Catarina Moreira and Andreas Wichert, Expert Systems with Applications, No. 40, pg. 5740-5754, 2013
CBIR with a Subspacee-tree: Principal Component Analysis versus Averaging
pdf iconbib icon
Andreas Wichert and Andre Verissimo, Multimedia Systems, Vol. 18, No. 4, pg. 83-293, 2012
Iterative random projections for high-dimensional data clustering
pdf iconbib icon
Ângelo Cardoso and Andreas Wichert, Pattern Recognition Letters, Vol. 33, No. 13, pg. 1749 - 1755, 2012
Product quantization for vector retrieval with no error
pdf iconbib icon
Andreas Wichert, Proceeedings of International Conference on Enterprise Information Systems, Vol. 1, pg. 87-92, 2012