80 Million Tiny Images: A Large Data Set for Nonparametric Object and Scene Recognition

Torralba, Antonio; Fergus, Rob; Freeman, William T.

doi:10.1109/tpami.2008.128

Cited by 1,724 publications

(1,085 citation statements)

References 36 publications

Supporting

Mentioning

1,062

Contrasting

Unclassified

Order By: Relevance

“…of Arxiv General Relativity) has 5.242 nodes and 28.980 edges whereas the largest graph (Amazon product copurchasing network from June 1) has 403, 394 nodes and 3, 387, 388 edges. To demonstrate the principal suitability of LS-DCUR for processing very large, dense matrices, we consider a data set of 80 million Google TM images [29]. Viewing each image (represented as a vector of gist descriptors) as a column, this results in a gigantic matrix containing 3072 · 10 7 elements.…”

Section: Methodsmentioning

confidence: 99%

“…When dealing with today's large-scale data sets, many data mining practitioners therefore often abandon deterministic approaches and resort to randomized approaches. However, huge data such as collections of online books at Amazon TM , image repositories at Flickr TM or Google TM , or personal health records [16,19,23,24,29] are becoming ever more common and thus pose a challenge to research on interpretable matrix factorization.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deterministic CUR for Improved Large-Scale Data Analysis: An Empirical Study

Thurau¹,

Kersting

Bauckhage³

2012

Proceedings of the 2012 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

Low-rank approximations which are computed from selected rows and columns of a given data matrix have attracted considerable attention lately. They have been proposed as an alternative to the SVD because they naturally lead to interpretable decompositions which was shown to be successful in application such as fraud detection, fMRI segmentation, and collaborative filtering. The CUR decomposition of large matrices, for example, samples rows and columns according to a probability distribution that depends on the Euclidean norm of rows or columns or on other measures of statistical leverage. At the same time, there are various deterministic approaches that do not resort to sampling and were found to often yield factorization of superior quality with respect to reconstruction accuracy. However, these are hardly applicable to large matrices as they typically suffer from high computational costs. Consequently, many practitioners in the field of data mining have abandon deterministic approaches in favor of randomized ones when dealing with today's large-scale data sets. In this paper, we empirically disprove this prejudice. We do so by introducing a novel, linear-time, deterministic CUR approach that adopts the recently introduced Simplex Volume Maximization approach for column selection. The latter has already been proven to be successful for NMF-like decompositions of matrices of billions of entries. Our exhaustive empirical study on more than 30 synthetic and real-world data sets demonstrates that it is also beneficial for CUR-like decompositions. Compared to other deterministic CUR-like methods, it provides comparable reconstruction quality but operates much faster so that it easily scales to matrices of billions of elements. Compared to sampling-based methods, it provides competitive reconstruction quality while staying in the same run-time complexity class.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Deterministic CUR for Improved Large-Scale Data Analysis: An Empirical Study

Thurau¹,

Kersting

Bauckhage³

2012

Proceedings of the 2012 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

show abstract

“…Li et al [25] presented a topdown saliency approach to incorporate low-level features and the objectness measure via label propagation. Generally, such task-driven methods are useful especially for object recognition [7], but they require knowledge learning that increases the complexity of saliency detection in general.…”

Section: A Top-down Saliency Detectionmentioning

confidence: 99%

“…From the perspective of information processing, those saliency algorithms can be broadly categorized as either top-down or bottom-up approaches. Top-down approaches [7]- [10] are goal-directed and usually adopt supervised learning with a specific class. Most of the saliency detection methods are based on bottomup visual attention mechanisms [11]- [15], [17], [18], [21], which are independent of the knowledge of the content in the image and utilize various low level features, such as intensity, color and orientation.…”

Section: Introductionmentioning

confidence: 99%

Correspondence Driven Saliency Transfer

Yang

Shen

Shao

et al. 2016

IEEE Trans. on Image Process.

138

View full text Add to dashboard Cite

Abstract-In this paper, we show that large annotated data sets have great potential to provide strong priors for saliency estimation rather than merely serving for benchmark evaluations. To this end, we present a novel image saliency detection method called saliency transfer. Given an input image, we first retrieve a support set of best matches from the large database of saliency annotated images. Then, we assign the transitional saliency scores by warping the support set annotations onto the input image according to computed dense correspondences. To incorporate context, we employ two complementary correspondence strategies: a global matching scheme based on scene-level analysis and a local matching scheme based on patch-level inference. We then introduce two refinement measures to further refine the saliency maps and apply the random-walk-with-restart by exploring the global saliency structure to estimate the affinity between foreground and background assignments. Extensive experimental results on four publicly available benchmark data sets demonstrate that the proposed saliency algorithm consistently outperforms the current state-of-the-art methods.

show abstract

“…is generated from the TinyImages collection [24], which provides 80 million color images of size 32 × 32. For each image, we first convert it to grayscale, compute its two-dimensional DCT (Discrete Cosine Transform), and then keep only the top 2% largest coefficients in magnitude.…”

Section: Numerical Experimentsmentioning

confidence: 99%

LSRN: A Parallel Iterative Solver for Strongly Over- or Underdetermined Systems

Meng

Saunders

Mahoney

2014

SIAM J. Sci. Comput.

112

127

View full text Add to dashboard Cite

We describe a parallel iterative least squares solver named that is based on random normal projection. computes the min-length solution to minx∈ℝn ‖Ax − b‖2, where A ∈ ℝm × n with m ≫ n or m ≪ n, and where A may be rank-deficient. Tikhonov regularization may also be included. Since A is involved only in matrix-matrix and matrix-vector multiplications, it can be a dense or sparse matrix or a linear operator, and automatically speeds up when A is sparse or a fast linear operator. The preconditioning phase consists of a random normal projection, which is embarrassingly parallel, and a singular value decomposition of size ⌈γ min(m, n)⌉ × min(m, n), where γ is moderately larger than 1, e.g., γ = 2. We prove that the preconditioned system is well-conditioned, with a strong concentration result on the extreme singular values, and hence that the number of iterations is fully predictable when we apply LSQR or the Chebyshev semi-iterative method. As we demonstrate, the Chebyshev method is particularly efficient for solving large problems on clusters with high communication cost. Numerical results show that on a shared-memory machine, is very competitive with LAPACK’s DGELSD and a fast randomized least squares solver called Blendenpik on large dense problems, and it outperforms the least squares solver from SuiteSparseQR on sparse problems without sparsity patterns that can be exploited to reduce fill-in. Further experiments show that scales well on an Amazon Elastic Compute Cloud cluster.

show abstract

80 Million Tiny Images: A Large Data Set for Nonparametric Object and Scene Recognition

Abstract: Abstract-

Cited by 1,724 publications

References 36 publications

Deterministic CUR for Improved Large-Scale Data Analysis: An Empirical Study

Deterministic CUR for Improved Large-Scale Data Analysis: An Empirical Study

Correspondence Driven Saliency Transfer

LSRN: A Parallel Iterative Solver for Strongly Over- or Underdetermined Systems

Contact Info

Product

Resources

About