Learning to Hash with Binary Reconstructive Embeddings

By Brian Kulis, Trevor Darrell. Neural Information Processing Systems, 2009

893 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/8c054e1f378e033f39a7b95c7f68da4b3b5544e4

How researchers see this paper

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

Citations in the corpus

Deep Supervised Hashing for Fast Image Retrieval (2016, methodology): “Specifically, as a representative family of such methods, many hashing algorithms (Gionis et al. 1999; Gong and Lazebnik 2011; Jiang and Li 2017; Kang et al. 2016; Kulis and Darrell 2009; Lai et al. 2015; Li et al. 2016; Lin et al. 2015a; Liu et al. 2017, 2014, 2012; Norouzi and Fleet 2011; Rastegar”
Hashing for Similarity Search: A Survey (2014, methodology): “Instead of relaxing the sign function to a continuous function, an alternative optimization scheme is presented in [63]: fixing all but one weight wmt and optimizing the problem 51 with respect to wmt.”
Image Matching from Handcrafted to Deep Features: A Survey (2020, methodology): “In this case, a plethora of methods have been proposed (Kulis and Darrell 2009; Wang et al. 2010; Strecha et al. 2012; Liu et al. 2012a; Norouzi and Blei 2011; Gong et al. 2013; Shakhnarovich 2005), among which image matching is considered an important experimental validation task.”
The Power of Asymmetry in Binary Hashing (2013, methodology): “Similar to [8, 5], we also constructed a synthetic 10D Uniform dataset, containing uniformly sampled 4000 points for a 10D hypercube.”
Semi-Supervised Hashing for Large-Scale Search (2012, background): “Based on the above formulation, a method called Binary Reconstructive Embedding (BRE) was designed to minimize a cost function measuring the difference between the metric and reconstructed distance in Hamming space.”
Hash Embeddings for Efficient Word Representations (2017, methodology): “However, since a hash function has a discrete codomain, it is not easy to optimize using e.g. gradient based methods used for training neural networks (Kulis and Darrell, 2009).”
Large-Scale Supervised Multimodal Hashing with Semantic Correlation Maximization (2014, methodology): “That is to say, most existing hashing methods are unimodal (Weiss, Torralba, and Fergus 2008; Andoni and Indyk 2008; Kulis and Darrell 2009; Lin, Ross, and Yagnik 2010; Wang, Kumar, and Chang 2010; Norouzi and Fleet 2011; Kong, Li, and Guo 2012; Heo et al. 2012; Liu et al. 2012a; Kong and Li 2012b; ”
Learning in high-dimensional multimedia data: the state of the art (2015, background): “64 bits 128 bits BRE [44] High low SSH [92] low low ITQ [23] fair fair KSH [53] High fair FastHash [50] High There are also some further developments extended from the standard hashing algorithms: Multiple Feature Hashing (MFH) [81] learns a group of hash functions to generate binary codes for multi”
HashNet: Deep Learning to Hash by Continuation (2017, methodology): “Binary Reconstruction Embedding (BRE) [19] pursues hash functions by minimizing the squared errors between the distances of data points and the distances of their corresponding hash codes.”
Fast Supervised Hashing with Decision Trees for High-Dimensional Data (2014, methodology): “Especially for highdimensional data, our method is orders of magnitude faster than many methods in terms of training time.”
Bit-Scalable Deep Hashing With Regularized Similarity Learning for Image Retrieval and Person Re-Identification (2015, methodology): “We compare with three person re-identiﬁcation methods (KISSME [43], eSDC [44], and FPNN [38]), four state-of-the-art hashing learning methods (BRE [39], MLH [12], KSH [7] and DRSH [40]) and the Euclidean distance.”
Multi-Task Learning with Low Rank Attribute Embedding for Person Re-Identification (2015, background): “In addition, each sample has a binary attribute vector ai ∈ {0, 1} , which may be semantic and labeled by humans or correspond to learned binary codes such as [22].”
Structure-Preserving Binary Representations for RGB-D Action Recognition (2016, methodology): “In our experiments, we compare the proposed method against seven general hashing algorithms including KSH [58], BRE [60], MLH [61], LSH [56], SpH [57], AGH [68], PCAH [69], BSSC [53] and RBM [70].”
Linear cross-modal hashing for efficient multimedia search (2013, methodology): “For example, supervised hashing methods [14, 17, 19, 24, 28] improve the hashing performance using the pre-provided pairs of data with the assumption that there is “similar” or “dissimilar” pairs in datasets.”
Learning to Search Efficiently in High Dimensions (2011, background): “Motivated by the success of machine learning in the design of ranking functions for information retrieval (the learning to rank problem [13, 9]) and the design of compact embedding into binary codes (the learning to hash problem [10]), it is natural to ask whether we can use machine learning (in par”
Inter-media hashing for large-scale retrieval from heterogeneous data sources (2013, background): “Binary reconstructive embedding [20] aims to align the hamming distance over the hash codes and the distance in the input Euclidean (and kernel) spaces.”
How should we evaluate supervised hashing? (2016, background): “Binary Reconstructive Embeddings (BRE) [5] minimize the distortion between the distance matrix in the original space and the Hamming distances of the codes.”