Manhattan hashing for large-scale image retrieval

Kong, Weihao; Li, Wu-Jun; Guo, Minyi

doi:10.1145/2348283.2348293

Cited by 87 publications

(66 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To do so, we test against three publicly available image datasets: 22k Labelme consisting of 22,019 images represented as 512 dimensional Gist descriptors [8]; CIFAR-10 a dataset of 60,000 images represented as 512 dimensional Gist descriptors; and 100k TinyImages a collection consisting of 100,000 images, represented by 384 dimensional Gist descriptors, randomly sub-sampled from the original 80 million tiny images dataset. The datasets and and associated features are identical to that used in previous related work [5] [4]. This ensures that our results are directly comparable to previously published figures.…”

Section: Datasetsmentioning

confidence: 58%

“…In contrast, vanilla LSH thresholds at zero and assigns a single bit per hyperplane. Existing multiple-bit quantisation schemes seek to position the thresholds based solely on information within the projected space, for example by using a k-means clustering on each projected dimension [5] [4]. In this work we show that the affinity between the data points in the original space can be a valuable signal for optimal threshold positioning.…”

Section: Neighbourhood Preserving Quan-tisationmentioning

confidence: 86%

“…Bottom: NPQ with 3 bits per dimension and seven thresholds. This is the same bit encoding as in [5].…”

Section: Related Workmentioning

confidence: 99%

“…As argued by previous authors [5] [4], thresholding at zero in this manner is suboptimal given that the area of highest data density typically occurs in the region around zero. There is therefore a high chance that neighbouring points may fall on different sides of the threshold, and therefore be assigned different bits in our encoding.…”

Section: Neighbourhood Preserving Quan-tisationmentioning

confidence: 99%

See 3 more Smart Citations

Neighbourhood preserving quantisation for LSH

Moran

Lavrenko

Osborne

2013

Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval

View full text Add to dashboard Cite

We introduce a scheme for optimally allocating multiple bits per hyperplane for Locality Sensitive Hashing (LSH). Existing approaches binarise LSH projections by thresholding at zero yielding a single bit per dimension. We demonstrate that this is a sub-optimal bit allocation approach that can easily destroy the neighbourhood structure in the original feature space. Our proposed method, dubbed Neighbourhood Preserving Quantization (NPQ), assigns multiple bits per hyperplane based upon adaptively learned thresholds. NPQ exploits a pairwise affinity matrix to discretise each dimension such that nearest neighbours in the original feature space fall within the same quantisation thresholds and are therefore assigned identical bits. NPQ is not only applicable to LSH, but can also be applied to any low-dimensional projection scheme. Despite using half the number of hyperplanes, NPQ is shown to improve LSH-based retrieval accuracy by up to 65% compared to the state-of-the-art.

show abstract

Section: Datasetsmentioning

confidence: 58%

Section: Neighbourhood Preserving Quan-tisationmentioning

confidence: 86%

“…Bottom: NPQ with 3 bits per dimension and seven thresholds. This is the same bit encoding as in [5].…”

Section: Related Workmentioning

confidence: 99%

Section: Neighbourhood Preserving Quan-tisationmentioning

confidence: 99%

See 2 more Smart Citations

Neighbourhood preserving quantisation for LSH

Moran

Lavrenko

Osborne

2013

Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval

View full text Add to dashboard Cite

show abstract

“…Kulis et al have extended LSH functions to a learned metric [33], which can also be considered as a supervised method. Beside these methods, several other hashing methods have been proposed to address different aspects of the modelling and computation, including semantic hashing [34], random maximum margin hashing [35], Manhattan hashing [36], dual-bit quantization hashing [37], spherical hashing [38] and k-means hashing [39].…”

Section: Related Workmentioning

confidence: 99%

Data-Dependent Hashing Based on p-Stable Distribution

Bai

Yang

Zhou

et al. 2014

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Abstract-The p-stable distribution is traditionally used for data-independent hashing. In this paper, we describe how to perform data-dependent hashing based on p-stable distribution. We commence by formulating the Euclidean distance preserving property in terms of variance estimation. Based on this property, we develop a projection method which maps the original data to arbitrary dimensional vectors. Each projection vector is a linear combination of multiple random vectors subject to p-stable distribution, in which the weights for the linear combination are learned based on the training data. An orthogonal matrix is then learned data-dependently for minimizing the thresholding error in quantization. Combining the projection method and the orthogonal matrix, we develop an unsupervised hashing scheme which preserves the Euclidean distance. Compared with data-independent hashing methods, our method takes the data distribution into consideration and gives more accurate hashing results with compact hash codes. Different from many datadependent hashing methods, our method accommodates multiple hash tables and is not restricted by the number of hash functions. To extend our method to a supervised scenario, we incorporate a supervised label propagation scheme into the proposed projection method. This results in a supervised hashing scheme which preserves semantic similarity of data. Experimental results show that our methods have outperformed several state-of-the-art hashing approaches in both effectiveness and efficiency.

show abstract

Unsupervised Online Hashing with Multi-Bit Quantization

Weng

Zhu

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Manhattan hashing for large-scale image retrieval

Cited by 87 publications

References 32 publications

Neighbourhood preserving quantisation for LSH

Neighbourhood preserving quantisation for LSH

Data-Dependent Hashing Based on p-Stable Distribution

Unsupervised Online Hashing with Multi-Bit Quantization

Contact Info

Product

Resources

About