Hamming Distance Metric Learning

By Mohammad Norouzi, David J. Fleet, Ruslan Salakhutdinov. 2012

663 citations recorded by the source corpus.

Subfields: Representation & Self-Supervised Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Representation & Self-Supervised Learning

Citations in the corpus

Revisiting IM2GPS in the Deep Learning Era (2017, neutral): “[5, 19, 35].”
On Mutual Information Maximization for Representation Learning (2019, background): “Secondly, it elucidates the importance of appropriately choosing the negative samples, which is indeed a critical component in deep metric learning based on triplet losses (Norouzi et al., 2012; Schroff et al., 2015).”
Understanding How Dimension Reduction Tools Work: An Empirical Approach to Deciphering t-SNE, UMAP, TriMAP, and PaCMAP for Data Visualization (2020, methodology): “Nevertheless, techniques such as triplets are used for both embedding (Chechik et al., 2010; Norouzi et al., 2012) and DR (Hadsell et al.”
Efficient Non-greedy Optimization of Decision Trees (2015, methodology): “[19, 20] used the same reformulation (5) of linear threshold functions to learn binary similarity preserving hash functions.”
A Survey on Metric Learning for Feature Vectors and Structured Data (2013, background): “A great advantage of working with binary codes is their small storage cost and the fact that exact neighbor search can be done in sublinear time (Norouzi et al., 2012b).”
Web-scale training for face identification (2014, methodology): “ also several ways to cope with the large dimensionality of the feature space. There are both unsupervised methods to compress features, like PCA or hashing based methods [18], and supervised methods [16]. It is commonly believed that compression of features leads to impoverished recognition accurac”
Metric Learning with Adaptive Density Discrimination (2015, background): “In this case, a ubiquitous solution is to enforce semantic similarity: examples of each class are demanded to be tightly clustered together, far from examples of other classes (for example, Schroff et al. (2015); Norouzi et al. (2012); Globerson & Roweis (2006); Chopra et al. (2005)).”
Deep Supervised Hashing for Fast Image Retrieval (2016, background): “Instead of constraining the precise numerical values of the distances, this loss function imposes constraints on the relative distances of all samples to the anchor sample Ia , and is thus considered more suitable for retrieval tasks (Norouzi et al. 2012).”
Simultaneous feature learning and hash coding with deep neural networks (2015, methodology): “Recently, some efforts [17, 10] have been made to learn hash functions that preserve relative similarities of the form “image I is more similar to image I than to image I−”.”
Improved Deep Metric Learning with Multi-class N-pair Loss Objective (2016, background): “Despite their wide use, both loss functions are known to suffer from slow convergence and they often require expensive data sampling method to provide nontrivial pairs or triplets to accelerate the training [2, 19, 17, 4].”
Central Similarity Quantization for Efficient Image and Video Retrieval (2019, methodology): “Most deep hashing methods [2, 48, 27, 18] learn hash functions by utilizing pairwise or triplet data similarity, where the data relationships are captured from a local perspective.”
Bit-Scalable Deep Hashing With Regularized Similarity Learning for Image Retrieval and Person Re-Identification (2015, background): “On the other hand, most existing hashing learning approaches generate the hashing codes with preset lengths ( e.g. , 16, 32 or 64 bits) [5][7][12], but one often requires hashing codes of different lengths under different scenarios.”
Learning Compact Binary Face Descriptor for Face Recognition (2015, background): “Recently, there have also been some works on feature learning-based (LE) face representation, and some of them have achieved reasonably good performance in face recognition.”
Discrete Graph Hashing (2014, methodology): “, Semantic Hashing [30], Binary Reconstruction Embeddings [16], Minimal Loss Hashing [27], Kernel-based Supervised Hashing [22], Hamming Distance Metric Learning [28], and Column Generation Hashing [20].”
Clustering-Oriented Representation Learning with Attractive-Repulsive Loss (2018, methodology): “This includes pairwise-based methods (Hsu and Kira 2016; Fogel et al. 2018; Hsu, Lv, and Kira 2018), and triplet-based methods (Norouzi, Fleet, and Salakhutdinov 2012; Schroff, Kalenichenko, and Philbin 2015; He et al. 2018); all of these require sophisticated methods for sampling training data,…”
Online multimodal deep similarity learning with application to image retrieval (2013, background): “…active research efforts in designing various distance/similarity measures on some low-level features using machine learning techniques, among which some works focus on learning to hashing or compact codes [39, 34, 20, 7, 23], and some others can be categorized into distance metric learning (DML).”
A Decade Survey of Content Based Image Retrieval Using Deep Learning (2020, background): “[26], Semi-supervised distance metric learning [27], Hamming distance metric learning [28], [29], and Rank based metric learning [30], [31].”
Fast Exact Search in Hamming Space With Multi-Index Hashing (2013, neutral): “Others focus on semantic similarity ( e.g., [25], [32], [31], [35], [26], [30], [21]).”
Learning Visual Semantic Relationships for Efficient Visual Retrieval (2015, methodology): “We introduce these techniques in the following part.”
Supervised Discrete Hashing (2015, methodology): “For BRE, MLH and KSH, we sample 5,000 images for training.”