Distributed Fast Supervised Discrete Hashing

Liu, Zhifeng; Feng, Chen; Duan, Shukai

doi:10.1109/access.2019.2924996

Cited by 6 publications

(8 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The experiment is evaluated the average precision and mean average precision (MAP) of each algorithm on dataset CIFAR-10. On dataset Wiki 2 , we compare the performance of three distributed supervision algorithms DSDHR, DSDH [36], and DFSDH [40]. DSDH regresses the hash code to the label information, while DFSDH regresses the label information to the hash code.…”

Section: B Analysis Of Computational Complexitymentioning

confidence: 99%

Distributed Supervised Discrete Hashing With Relaxation

et al. 2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

The data-dependent hash methods are becoming more and more attractive because they perform well in fast retrieval and storing high-dimensional data. Most existing supervised hashes are centralized, such as supervised discrete hashing (SDH) and supervised discrete hashing with relaxation (SDHR). The SDH algorithm determines the regression target by using ordinary least squares regression and the traditional zero-one matrix encoding of class label information. And SDHR is a constraint to the regression target matrix so that each example is correctly classified and satisfies a larger margin, so as to achieve the purpose of optimizing the regression target. In real environment, a large amount of data will be distributed in different nodes. Therefore, the centralized hash method has great limitations. In this article, we propose distributed supervised discrete hashing algorithm with relaxation (DSDHR) based on SDHR. The SDHR algorithm is introduced into the distributed network. In this framework, all nodes share a centralized hash learning model. At the same time, in order to ensure that the distributed hash algorithm is updated in parallel on multiple nodes, consistency constraints are introduced. In each node, alternative iterative methods are used to obtain the binary hash code, regression target and hash function. Experiments show that DSDHR has certain competitive advantages over some centralized hashing algorithms and some distributed hashing algorithms.INDEX TERMS Distributed optimization, Distributed hashing, Discrete hashing, Supervised learning.

show abstract

Section: B Analysis Of Computational Complexitymentioning

confidence: 99%

Distributed Supervised Discrete Hashing With Relaxation

et al. 2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

“…where W (l) ∈ R d×k denote the top PCA [40] vectors corresponding to the l-th individual, and k is the length of short codes, and tr() is the matrix trace, and I k is a tby-t identity matrix. The final projection matrix W = W (1) , W (2) , . .…”

Section: Bootstrapmentioning

confidence: 99%

“…In the machine learning, nearest neighbor search [1], [2] is one of the widely used technologies, and it has been applied in many areas, e.g. , data retrieval [3], [4], scene text recognition [5] and computer vision [6].…”

Section: Introductionmentioning

confidence: 99%

Unsupervised Ensemble Hashing: Boosting Minimum Hamming Distance

et al. 2020

View full text Add to dashboard Cite

Hashing aims at learning discriminative binary codes of high-dimensional data for the approximate nearest neighbor searching. However, the distance ranking obtained by traditional methods is not optimum in the Hamming space, and it degrades the performance for retrieval tasks. To tackle the above problem, an unsupervised ensemble hashing is proposed to improve the ranking accuracy by assembling the diverse hash tables independently in this paper. We observe that the higher the accuracy is the larger diversity the base learner has, and the more effective the ensemble method is. Based on this principle, two special ensembles hashing approaches are proposed to increase diversity by bootstrap sampling with data-dependent methods. Especially, the results are better when the minimum Hamming distance is large and the variance of the Hamming distance is small. This proposed method is conducted in the experiments and the results show that it can achieve about 10%-25% performance compared with the baseline algorithm, which achieves competitive results with the state-of-the-art methods on the CIFAR-10 and LabelMe benchmarks. INDEX TERMS Unsupervised hashing, ensemble method, accuracy and diversity, hamming distance, distance variance.

show abstract

“…Note that there exist a few distributed hashing algorithms [50,89,105,60] which is optimized by Alternating Direction Method of Multipliers (ADMM) [7]. To update the local model with ADMM, each local machine communicates with all adjacent machines to aggregate their latest local models.…”

Section: Approaches and Contributionsmentioning

confidence: 99%

“…By far, a few distributed hashing algorithms [50,89,105,60] The first distributed hashing algorithm was proposed by Leng et al [50], which minimizes the quantization error splitting with consensus constraints by ADMM. The followup works [60,89,105] were proposed to split the original objective function in [56,72] with consensus constraints and optimize the new loss function by ADMM.…”

Section: Distributed Hashingmentioning

confidence: 99%

Advanced topics in learning to hash for large-scale retrieval

Huang¹

View full text Add to dashboard Cite

show abstract

Distributed Fast Supervised Discrete Hashing

Cited by 6 publications

References 17 publications

Distributed Supervised Discrete Hashing With Relaxation

Distributed Supervised Discrete Hashing With Relaxation

Unsupervised Ensemble Hashing: Boosting Minimum Hamming Distance

Advanced topics in learning to hash for large-scale retrieval

Contact Info

Product

Resources

About