Home
Carlos Santos's Library tagged → View Popular
Clustering Billions of Images with Large Scale Nearest Neighbor Search
Image collections on this scale make performing even the most common and simple computer vision, image processing, and machine learning tasks non-trivial. An example is nearest neighbor search, which not only serves as a fundamental subproblem in many more sophisticated algorithms, but also has direct applications, such as image retrieval and image clustering. In this paper, we address the nearest neighbor problem as the first step towards scalable image processing. We describe a scalable version of an approximate nearest neighbor search algorithm and discuss how it can be used to find near duplicates among over a billion images.
-
Image collections on this scale make performing even the most common and simple computer vision, image processing, and machine learning tasks non-trivial. An example is nearest neighbor search, which not only serves as a fundamental subproblem in many more sophisticated algorithms, but also has direct applications, such as image retrieval and image clustering. In this paper, we address the nearest neighbor problem as the first step towards scalable image processing. We describe a scalable version of an approximate nearest neighbor search algorithm and discuss how it can be used to find near duplicates among over a billion images.
Cover Tree
A Cover Tree is a datastructure helpful in calculating the nearest neighbor of points given only a metric. A cover tree is particularly motivating for a confluence of reasons:
1. The running time of a nearest neighbor query is only O(log(n)) given a fixed intrinsic dimensionality. (like KR2002 and KL04)
2. The space usage and query time are O(n) under no assumptions. (like the naive approach, sb(s), and ball trees)
3. It's remarkably fast in practice.
-
- The running time of a nearest neighbor query is only O(log(n)) given a fixed intrinsic dimensionality. (like KR2002 and KL04)
- The space usage and query time are O(n) under no assumptions. (like the naive approach, sb(s), and ball trees)
- It's remarkably fast in practice.
A Cover Tree is a datastructure helpful in calculating the nearest
neighbor of points given only a metric. A cover tree is particularly
motivating for a confluence of reasons:
Selected Tags
Related Tags
Sponsored Links
Highlighter, Sticky notes, Tagging, Groups and Network: integrated suite dramatically boosting research productivity. Learn more »
Join Diigo