Maximum-likelihood approximate nearest neighbor method in real-time image recognition
From MaRDI portal
Abstract: We explore the problems of classification of composite object (images, speech signals) with low number of models per class. We study the question of improving recognition performance for medium-sized database (thousands of classes). The key issue of fast approximate nearest-neighbor methods widely applied in this task is their heuristic nature. It is possible to strongly prove their efficiency by using the theory of algorithms only for simple similarity measures and artificially generated tasks. On the contrary, in this paper we propose an alternative, statistically optimal greedy algorithm. At each step of this algorithm joint density (likelihood) of distances to previously checked models is estimated for each class. The next model to check is selected from the class with the maximal likelihood. The latter is estimated based on the asymptotic properties of the Kullback-Leibler information discrimination and mathematical model of piecewise-regular object with distribution of each regular segment of exponential type. Experimental results in face recognition for FERET dataset prove that the proposed method is much more effective than not only brute force and the baseline (directed enumeration method) but also approximate nearest neighbor methods from FLANN and NonMetricSpaceLib libraries (randomized kd-tree, composite index, perm-sort).
Recommendations
- The maximal likelihood enumeration method for the problem of classifying piecewise regular objects
- Clustering and maximum likelihood search for efficient statistical classification with medium-sized databases
- A fast nearest neighbor search algorithm by filtration
- scientific article; zbMATH DE number 1759582
- Fast and versatile algorithm for nearest neighbor search based on a lower bound tree
Cites work
- An optimal algorithm for approximate nearest neighbor searching fixed dimensions
- Clustering to minimize the maximum intercluster distance
- Computer vision. Models, learning, and inference. Foreword by Andrew Fitzgibbon.
- Face recognition from a single image per person: a survey
- scientific article; zbMATH DE number 859030 (Why is no real title available?)
- scientific article; zbMATH DE number 3241743 (Why is no real title available?)
- Multidimensional binary search trees used for associative searching
- Pivot selection techniques for proximity searching in metric spaces
- Probabilistic neural network with homogeneity testing in recognition of discrete patterns set
- Probabilistic proximity search: Fighting the curse of dimensionality in metric spaces.
- Search techniques in intelligent classification systems
Cited in
(5)- Sequential three-way decisions in multi-category image recognition with deep features based on distance factor
- The maximal likelihood enumeration method for the problem of classifying piecewise regular objects
- scientific article; zbMATH DE number 1759582 (Why is no real title available?)
- Robust Image Recognition Algorithm of Maximum Likelihood Estimation Sparse Representation Based on Class-related Neighbors Subspace
- Clustering and maximum likelihood search for efficient statistical classification with medium-sized databases
This page was built for publication: Maximum-likelihood approximate nearest neighbor method in real-time image recognition
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2289612)
