Results 1  10
of
32
Fast and Robust Earth Mover’s Distances
"... We present a new algorithm for a robust family of Earth Mover’s Distances EMDs with thresholded ground distances. The algorithm transforms the flownetwork of the EMD so that the number of edges is reduced by an order of magnitude. As a result, we compute the EMD by an order of magnitude faster tha ..."
Abstract

Cited by 90 (6 self)
 Add to MetaCart
(Show Context)
We present a new algorithm for a robust family of Earth Mover’s Distances EMDs with thresholded ground distances. The algorithm transforms the flownetwork of the EMD so that the number of edges is reduced by an order of magnitude. As a result, we compute the EMD by an order of magnitude faster than the original algorithm, which makes it possible to compute the EMD on large histograms and databases. In addition, we show that EMDs with thresholded ground distances have many desirable properties. First, they correspond to the way humans perceive distances. Second, they are robust to outlier noise and quantization effects. Third, they are metrics. Finally, experimental results on image retrieval show that thresholding the ground distance of the EMD improves both accuracy and speed. 1.
Modeling LSH for Performance Tuning
"... Although LocalitySensitive Hashing (LSH) is a promising approach to similarity search in highdimensional spaces, it has not been considered practical partly because its search quality is sensitive to several parameters that are quite data dependent. Previous research on LSH, though obtained intere ..."
Abstract

Cited by 42 (1 self)
 Add to MetaCart
Although LocalitySensitive Hashing (LSH) is a promising approach to similarity search in highdimensional spaces, it has not been considered practical partly because its search quality is sensitive to several parameters that are quite data dependent. Previous research on LSH, though obtained interesting asymptotic results, provides little guidance on how these parameters should be chosen, and tuning parameters for a given dataset remains a tedious process. To address this problem, we present a statistical performance model of Multiprobe LSH, a stateoftheart variance of LSH. Our model can accurately predict the average search quality and latency given a small sample dataset. Apart from automatic parameter tuning with the performance model, we also use the model to devise an adaptive LSH search algorithm to determine the probing parameter dynamically for each query. The adaptive probing method addresses the problem that even though the average performance is tuned for optimal, the variance of the performance is extremely high. We experimented with three different datasets including audio, images and 3D shapes to evaluate our methods. The results show the accuracy of the proposed model: the recall errors predicted are within 5 % from the real values for most cases; the adaptive search method reduces the standard deviation of recall by about 50 % over the existing method.
Efficient KNearest Neighbor Graph Construction for Generic Similarity Measures
"... KNearest Neighbor Graph (KNNG) construction is an important operation with many web related applications, including collaborative filtering, similarity search, and many others in data mining and machine learning. Existing methods for KNNG construction either do not scale, or are specific to certa ..."
Abstract

Cited by 27 (0 self)
 Add to MetaCart
(Show Context)
KNearest Neighbor Graph (KNNG) construction is an important operation with many web related applications, including collaborative filtering, similarity search, and many others in data mining and machine learning. Existing methods for KNNG construction either do not scale, or are specific to certain similarity measures. We present NNDescent, a simple yet efficient algorithm for approximate KNNG construction with arbitrary similarity measures. Our method is based on local search, has minimal space overhead and does not rely on any shared global index. Hence, it is especially suitable for largescale applications where data structures need to be distributed over the network. We have shown with a variety of datasets and similarity measures that the proposed method typically converges to above 90 % recall with each point comparing only to several percent of the whole dataset on average.
Earth Mover Distance over HighDimensional Spaces
, 2007
"... The Earth Mover Distance (EMD) between two equalsize sets of points in R d is defined to be the minimum cost of a bipartite matching between the two pointsets. It is a natural metric for comparing sets of features, and as such, it has received significant interest in computer vision. Motivated by re ..."
Abstract

Cited by 22 (8 self)
 Add to MetaCart
The Earth Mover Distance (EMD) between two equalsize sets of points in R d is defined to be the minimum cost of a bipartite matching between the two pointsets. It is a natural metric for comparing sets of features, and as such, it has received significant interest in computer vision. Motivated by recent developments in that area, we address computational problems involving EMD over highdimensional pointsets. A natural approach is to embed the EMD metric into ℓ1, and use the algorithms designed for the latter space. However, Khot and Naor [KN06] show that any embedding of EMD over the ddimensional Hamming cube into ℓ1 must incur a distortion Ω(d), thus practically losing all distance information. We circumvent this roadblock by focusing on sets with cardinalities upperbounded by a parameter s, and achieve a distortion of only O(log s · log d). Since in applications the feature sets have bounded size, the resulting distortion is much smaller than the Ω(d) lower bound. Our approach is quite general and easily extends to EMD over R d. We then provide a strong lower bound on the multiround communication complexity of estimating EMD, which in particular strengthens the known nonembeddability result of [KN06]. Our bound exhibits a smooth tradeoff between approximation and communication, and for example implies that every algorithm that estimates EMD using constant size sketches can only achieve Ω(log s) approximation.
Asymmetric distance estimation with sketches for similarity search in highdimensional spaces
 In SIGIR
, 2008
"... Efficient similarity search in highdimensional spaces is important to contentbased retrieval systems. Recent studies have shown that sketches can effectively approximate L1 distance in highdimensional spaces, and that filtering with sketches can speed up similarity search by an order of magnitude ..."
Abstract

Cited by 20 (0 self)
 Add to MetaCart
(Show Context)
Efficient similarity search in highdimensional spaces is important to contentbased retrieval systems. Recent studies have shown that sketches can effectively approximate L1 distance in highdimensional spaces, and that filtering with sketches can speed up similarity search by an order of magnitude. It is a challenge to further reduce the size of sketches, which are already compact, without compromising accuracy of distance estimation. This paper presents an efficient sketch algorithm for similarity search with L2 distances and a novel asymmetric distance estimation technique. Our new asymmetric estimator takes advantage of the original feature vector of the query to boost the distance estimation accuracy. We also apply this asymmetric method to existing sketches for cosine similarity and L1 distance. Evaluations with datasets extracted from images and telephone records show that our L2 sketch outperforms existing methods, and the asymmetric estimators consistently improve the accuracy of different sketch methods. To achieve the same search quality, asymmetric estimators can reduce the sketch size by 10 % to 40%.
Efficiently matching sets of features with random histograms
 in ACM Multimedia
, 2008
"... As the commonly used representation of a featurerich data object has evolved from a single feature vector to a set of feature vectors, a key challenge in building a contentbased search engine for featurerich data is to match featuresets efficiently. Although substantial progress has been made du ..."
Abstract

Cited by 18 (0 self)
 Add to MetaCart
(Show Context)
As the commonly used representation of a featurerich data object has evolved from a single feature vector to a set of feature vectors, a key challenge in building a contentbased search engine for featurerich data is to match featuresets efficiently. Although substantial progress has been made during the past few years, existing approaches are still inefficient and inflexible for building a search engine for massive datasets. This paper presents a randomized algorithm to embed a set of features into a single highdimensional vector to simplify the featureset matching problem. The main idea is to project feature vectors into an auxiliary space using locality sensitive hashing and to represent a set of features as a histogram in the auxiliary space. A histogram is simply a high dimensional vector, and efficient similarity measures like L1 and L2 distances can be employed to approximate featureset distance measures. We evaluated the proposed approach under three different task settings, i.e. contentbased image search, image object recognition and nearduplicate video clip detection. The experimental results show that the proposed approach is indeed effective and flexible. It can achieve accuracy comparable to the featureset matching methods, while requiring significantly less space and time. For object recognition with Caltech 101 dataset, our method runs 25 times faster to achieve the same precision as Pyramid Matching Kernel, the stateoftheart featureset matching method.
Matching Point Sets with respect to the Earth Mover’s Distance
 Proc. 13th Annu. Euro. Symp. Algo
, 2005
"... Let A and B be two sets of m resp. n weighted points in the plane, with m ≤ n. We present (1 + ɛ) and (2+ɛ)approximation algorithms for the minimum Euclidean Earth Mover’s Distance between A and B under translations and rigid motions respectively. In the general case where the sets have unequal tot ..."
Abstract

Cited by 14 (2 self)
 Add to MetaCart
(Show Context)
Let A and B be two sets of m resp. n weighted points in the plane, with m ≤ n. We present (1 + ɛ) and (2+ɛ)approximation algorithms for the minimum Euclidean Earth Mover’s Distance between A and B under translations and rigid motions respectively. In the general case where the sets have unequal total weights the algorithms run in O((n 3 m/ɛ 4) log 2 (n/ɛ)) time for translations and O((n 4 m 2 /ɛ 4) log 2 (n/ɛ)) time for rigid motions. When the sets have equal total weights, the respective running times decrease to O((n 2 /ɛ 4) log 2 (n/ɛ)) and O((n 3 m/ɛ 4) log 2 (n/ɛ)). We also show how to compute a (1 + ɛ) and (2 + ɛ)approximation of the minimum cost Euclidean bipartite matching under translations and rigid motions in O((n 3/2 /ɛ 7/2) log 5 n)andO((n/ɛ) 7/2 log 5 n) time respectively. 1
NPIC: Hierarchical synthetic image classification using image search and generic features", CIVR’06
 Proc. of Conf. on Image and Video Retrieval
, 2006
"... Abstract. We introduce NPIC, an image classification system that focuses on synthetic (e.g., nonphotographic) images. We use classspecific keywords in an image search engine to create a noisily labeled training corpus of images for each class. NPIC then extracts both contentbased image retrieval ..."
Abstract

Cited by 9 (1 self)
 Add to MetaCart
(Show Context)
Abstract. We introduce NPIC, an image classification system that focuses on synthetic (e.g., nonphotographic) images. We use classspecific keywords in an image search engine to create a noisily labeled training corpus of images for each class. NPIC then extracts both contentbased image retrieval (CBIR) features and metadatabased textual features for each image for machine learning. We evaluate this approach on three different granularities: 1) natural vs. synthetic, 2) map vs. figure vs. icon vs. cartoon vs. artwork 3) and further subclasses of the map and figure classes. The NPIC framework achieves solid performance (99%, 97% and 85 % in cross validation, respectively). We find that visual features provide a significant boost in performance, and that textual and visual features vary in usefulness at the different levels of granularities of classification. 1
iscope: personalized multimodality image search for mobile devices
 In Proceedings of Mobisys ’09
, 2009
"... Mobile devices are becoming a primary medium for personal information gathering, management, and sharing. Managing personal image data on mobile platforms is a difficult problem due to large data set size, content diversity, heterogeneous individual usage patterns, and resource constraints. This art ..."
Abstract

Cited by 8 (0 self)
 Add to MetaCart
(Show Context)
Mobile devices are becoming a primary medium for personal information gathering, management, and sharing. Managing personal image data on mobile platforms is a difficult problem due to large data set size, content diversity, heterogeneous individual usage patterns, and resource constraints. This article presents a usercentric system, called iScope, for personal image management and sharing on mobile devices. iScope uses multimodality clustering of both content and context information for efficient image management and search, and online learning techniques for predicting images of interest. It also supports distributed contentbased search among networked devices while maintaining the same intuitive interface, enabling efficient information sharing among people. We have implemented iScope and conducted infield experiments using networked Nokia N810 portable Internet tablets. Energy efficiency was a primary design focus during the design and implementation of the iScope search algorithms. Experimental results indicate that iScope improves search time and search energy by 4.1 × and 3.8 × on average, relative to browsing.
Tradeoffs in Approximate Range Searching Made Simpler
 XXI BRAZILIAN SYMPOSIUM ON COMPUTER GRAPHICS AND IMAGE PROCESSING
, 2008
"... ..."
(Show Context)