Scalable Graph Hashing with Feature Transformation Motivation (Unsupervised)Graph Hashing o Guide the hashing code learning procedure by directly exploiting the pairwise similarity (neighborhood structure). min >Sigllbi-bill2=tr(BTCB) subject to:biE{-1,1c ∑b:=0 h∑bb=I oShould be expected to achieve better performance than non-graph based methods if the learning algorithms are effective. 口卡，24元互Q0 Li (http://cs.nju.edu.cn/lwj) Learning to Hash LAMDA,CS.NJU 19/43Scalable Graph Hashing with Feature Transformation Motivation (Unsupervised) Graph Hashing Guide the hashing code learning procedure by directly exploiting the pairwise similarity (neighborhood structure). minX ij Sij ||bi − bj ||2 = tr(B TLB) subject to : bi ∈ {−1, 1} c X i bi = 0 1 n X i bib T i = I Should be expected to achieve better performance than non-graph based methods if the learning algorithms are effective. Li (http://cs.nju.edu.cn/lwj) Learning to Hash LAMDA, CS, NJU 19 / 43