A General Two-Step Approach to Learning-Based Hashing
Citations
925 citations
Cites background or methods from "A General Two-Step Approach to Lear..."
...Similar to the two-step paradigm in TSH (Lin et al. 2013), the proposed method decomposes the learning process into a hash code learning step and a hash function learning step....
[...]
...Hence, similar to existing methods (Liu et al. 2012; Lin et al. 2013; Kulis and Darrell 2009), if in some iteration (of the outer loop) increases the value of (3), Algorithm 1 does not update H and L and continues the loop (see the IF-ELSE block in the outer loop of Algorithm 1)....
[...]
...Note that there exist other algorithms to factorize S into HH , such as the block coordinate descent (BCD) algorithm in (Lin et al. 2013)....
[...]
...Hence, similar to existing methods (Liu et al. 2012; Lin et al. 2013; Kulis and Darrell 2009), if in some iteration (of the...
[...]
...To avoid this issue, one of the popular ways is decomposing the learning process into a hash code learning stage followed by a hash function learning stage (e.g., (Zhang et al. 2010; Lin et al. 2013))....
[...]
923 citations
Cites methods from "A General Two-Step Approach to Lear..."
...The similar idea was also adopted in [17,18,39]....
[...]
838 citations
807 citations
531 citations
Cites methods from "A General Two-Step Approach to Lear..."
...The paper [77] presents a general two-step approach to learning-based hashing: learn binary embedding (codes) and then learn the hash function mapping the input item to the learnt binary codes....
[...]
References
3,705 citations
Additional excerpts
...We compare with a few state-of-the-art hashing methods, including 6 (semi-)supervised methods: Supervised Hashing with Kernels (KSH) [8], Iterative Quantization with supervised embedding (ITQ-CCA) [3], Minimal Loss Hashing (MLH) [10], Supervised Binary Reconstructive Embeddings (BREs) [5] and its unsupervised version BRE, Supervised Self-Taught Hashing (STHs) [16] and its unsupervised version STH, Semi-supervised sequential Projection Learning Hashing(SPLH) [13], and 7 unsupervised methods: Locality-Sensitive Hashing (LSH) [2], Iterative Quantization (ITQ) [3], Anchor Graph Hashing (AGH) [9], Spectral Hashing (SPH [15]), Spherical Hashing (SPHER) [4], Multi-dimension Spectral Hashing (MDSH) [14], Kernelized Locality-Sensitive Hashing KLSH [6]....
[...]
...The common forms of hash function include linear perceptron functions (MLH, SPLH, LSH), kernel functions (KSH, KLSH), eigenfunctions (SH, MDSH)....
[...]
...One of the seminal approaches in this vain is Locality-Sensitive Hashing (LSH) [2], which randomly generates hash functions to approximate cosine similarity....
[...]
...For each dataset, the bandwidth parameters of Gaussian affinity in MDSH and RBF kernel in KLSH, KSH and our method TSH is set as σ = td̄....
[...]
2,776 citations
[...]
2,641 citations
1,871 citations
"A General Two-Step Approach to Lear..." refers background in this paper
...Applications in computer vision include content-based image retrieval, object recognition [12], image matching, etc....
[...]
1,697 citations