Locality-sensitive hashing

About: Locality-sensitive hashing is a research topic. Over the lifetime, 1894 publications have been published within this topic receiving 69362 citations.

Event detection and trending in multiple social networking sites

Abstract: A continuous rise in popularity of social media motivates many people to express their opinions and news on the real-time basis. In this paper, the social networking sites such as Twitter and Facebook are considered as a platform for event detection. Since social information streams are sparse and continuous, the processing time and speed are vital while detecting events. We suggest a novel approach of discovering events from multiple social streams using widely used Euclidean realization of locality sensitive hashing (LSH) algorithm. In our proposed method, the LSH is used twice in event detection. Firstly, it is used to obtain the events independently from both social streams. The cross-over events between social networks are detected by applying the algorithm one more time. The detected events can be trended to show their activeness on different networks. We explore a theoretical approach on the design of event detection and trending in multiple social sites.

Supervised Deep Hashing for Hierarchical Labeled Data

Abstract: Recently, hashing methods have been widely used in large-scale image retrieval. However, most existing hashing methods did not consider the hierarchical relation of labels, which means that they ignored the rich information stored in the hierarchy. Moreover, most of previous works treat each bit in a hash code equally, which does not meet the scenario of hierarchical labeled data. In this paper, we propose a novel deep hashing method, called supervised hierarchical deep hashing (SHDH), to perform hash code learning for hierarchical labeled data. Specifically, we define a novel similarity formula for hierarchical labeled data by weighting each layer, and design a deep convolutional neural network to obtain a hash code for each data point. Extensive experiments on several real-world public datasets show that the proposed method outperforms the state-of-the-art baselines in the image retrieval task.

Fast K-means for Large Scale Clustering

Abstract: K-means algorithm has been widely used in machine learning and data mining due to its simplicity and good performance. However, the standard k-means algorithm would be quite slow for clustering millions of data into thousands of or even tens of thousands of clusters. In this paper, we propose a fast k-means algorithm named multi-stage k-means (MKM) which uses a multi-stage filtering approach. The multi-stage filtering approach greatly accelerates the k-means algorithm via a coarse-to-fine search strategy. To further speed up the algorithm, hashing is introduced to accelerate the assignment step which is the most time-consuming part in k-means. Extensive experiments on several massive datasets show that the proposed algorithm can obtain up to 600X speed-up over the k-means algorithm with comparable accuracy.

Fast t-SNE algorithm with forest of balanced LSH trees and hybrid computation of repulsive forces

Abstract: An acceleration of the well-known t-Stochastic Neighbor Embedding (t-SNE) (Hinton and Roweis, 2003; Maaten and Hinton, 2008) algorithm, probably the best (nonlinear) dimensionality reduction and visualization method, is proposed in this article. By using a specially-tuned forest of balanced trees constructed via locality sensitive hashing is improved significantly upon the results presented in Maaten (2014), achieving a complexity significantly closer to true O ( n log n ) , and vastly improving behavior for huge numbers of instances and attributes. Such acceleration removes the necessity to use PCA to reduce dimensionality before the start of t-SNE. Additionally, a fast hybrid method for repulsive forces computation (a part of the t-SNE algorithm), which is currently the fastest method known, is proposed. A parallelized version of our algorithm, characterized by a very good speedup factor, is proposed.

LSH for loop closing detection in underwater visual SLAM

Abstract: Effectiveness in loop closing detection is crucial to increase accuracy in SLAM (Simultaneous Localization and Mapping) for mobile robots. The most representative approaches to visual loop closing detection are based on feature matching or BOW (Bag of Words), being slow and needing a lot of memory resources or a previously defined vocabulary, which complicates and delays the whole process. This paper present a new visual LSH (Locality Sensitive Hashing)-based approach for loop closure detection, where images are hashed to accelerate considerably the whole comparison process. The algorithm is applied in AUV (Autonomous Underwater Vehicles), in several aquatic scenarios, showing promising results and the validity of this proposal to be applied online.