Minimal Loss Hashing for Compact Binary Codes

By Mohammad Norouzi, David J. Fleet. International Conference on Machine Learning, 2011

767 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/5ff8c048a042d37a99a9f9c5acb0d82972a45ad0

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

Feature Learning Based Deep Supervised Hashing with Pairwise Labels (2015, methodology): “Representative pairwise label based methods include sequential projection learning for hashing (SPLH) [Wang et al., 2010], minimal loss hashing (MLH) [Norouzi and Fleet, 2011], supervised hashing with kernels (KSH) [Liu et al., 2012], two-step hashing (TSH) [Lin et al., 2013], fast supervised hashin”
Optimized Product Quantization (2014, methodology): “Binary embedding is a popular way of encoding vectors [9], [15], [16], [17], [18], [19], [27].”
Deep Cauchy Hashing for Hamming Space Retrieval (2018, methodology): “Many learning to hash methods have been proposed to enable efficient ANN search by Hamming ranking of compact hash codes, including both supervised and unsupervised methods [16, 12, 26, 9, 22, 30, 25, 24, 35].”
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].”
A Survey on Learning to Hash (2016, methodology): “BRE = binary reconstructive embedding [66], MLH = minimal loss hashing [107], LSH = locality sensitive hashing [14], ITQ = iterative quantization [35], [36], SH = spectral hashing [156], AGH-2 = two-layer hashing with graphs [86], USPLH = unsupervised sequential projection learning hashing [143], PQ”
Deep semantic ranking based hashing for multi-label image retrieval (2015, methodology): “To preserve semantic structure of the data, hashing methods with supervisory information in form of class labels have been further developed [24, 27, 28, 19, 17].”
DeepHash: Getting Regularization, Depth and Fine-Tuning Right (2015, background): “…are Iterative Quantization [9], Spectral Hashing [10], Restricted Boltzmann Machines [8], while some examples of state-of-the-art supervised schemes include Minimal Loss Hashing [11], Kernel-based Supervised Hashing [12], Ranking-based Supervised Hashing [13] and Column Generation Hashing [14].”
Optimized Cartesian K-Means (2014, background): “Typical approaches are unsupervised hashing [5], [12], [13], [33], [36], [40], [41], [42] and supervised hashing [16], [23], with kernelized version [7], [17], and extensions to multimodality [31], [32], [43], etc.”
Supervised Hashing for Image Retrieval via Image Representation Learning (2014, methodology): “CNNH+ 0.969 0.975 0.971 0.975 0.465 0.521 0.521 0.532 0.623 0.630 0.629 0.625 CNNH 0.957 0.963 0.956 0.960 0.439 0.511 0.509 0.522 0.611 0.618 0.625 0.608 KSH 0.872 0.891 0.897 0.900 0.303 0.337 0.346 0.356 0.556 0.572 0.581 0.588 ITQ-CCA 0.659 0.694 0.714 0.726 0.264 0.282 0.288 0.295 0.435 0.435 0”
HashNet: Deep Learning to Hash by Continuation (2017, methodology): “Typical supervised methods include Binary Reconstruction Embedding (BRE) [19], Minimal Loss Hashing (MLH) [30] and Hamming Distance Metric Learning [31].”
Simultaneous feature learning and hash coding with deep neural networks (2015, methodology): “We test and compare the search accuracies of the proposed method with eight state-of-the-art hashing methods, including three unsupervised methods LSH [2], SH [26] and ITQ [4], and five supervised methods CNNH [27], KSH [12], MLH [16], BRE [8] and ITQ-CCA [4].”
Learning Discrete Representations via Information Maximizing Self-Augmented Training (2017, methodology): “The representative ones include K-means clustering and hashing (He et al., 2013), Gaussian mixture model clustering, iterative quantization (Gong et al., 2013), and minimal-loss hashing (Norouzi & Blei, 2011).”
Circulant Binary Embedding (2014, methodology): “Following (Gong et al., 2013a; Norouzi & Fleet, 2012; Gordo & Perronnin, 2011), we use 10,000 randomly sam- pled instances for training.”
Deep hashing for compact binary codes learning (2015, methodology): “Table 4 shows the training and test time of different hashing methods on the CIFAR-10 dataset when 16-bit binary codes are used for evaluation , where the test time is computed for each query 0 200 400 600 800 1000 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 N Recall ITQ KMH Spherical SH PCAH DH (a) 16 bits 0”