Introduction Multimodal Methods sunset lardseape grass hdr b带ch sky b带atmosph带rw fly white Back feathers sun direct green field sky wide hoy包ram sky Animal Multi-Source Hashing Cross-Modal Hashing 口卡得三4元互Q0 Li (http://cs.nju.edu.cn/lvj) Learning to Hash C5.NJU16/210
Introduction Multimodal Methods Multi-Source Hashing Cross-Modal Hashing Li (http://cs.nju.edu.cn/lwj) Learning to Hash CS, NJU 16 / 210
Introduction Multi-Source Hashing o Aim at learning better codes than unimodal hashing by leveraging auxiliary views. o Assume that all the views are provided for a query. MFH:Multiple feature hashing(Song et al.,2011) o CH:Composite hashing (Zhang et al.,2011) 日卡三4元，互Q0 Li (http://cs.nju.edu.cn/lwj) Learning to Hash CS.NJU 17 /210
Introduction Multi-Source Hashing Aim at learning better codes than unimodal hashing by leveraging auxiliary views. Assume that all the views are provided for a query. MFH: Multiple feature hashing (Song et al., 2011) CH: Composite hashing (Zhang et al., 2011) Li (http://cs.nju.edu.cn/lwj) Learning to Hash CS, NJU 17 / 210
Introduction Cross-Modal Hashing Given a query of either image or text,return images or texts similar to it. CVH:Cross view hashing(Kumar and Udupa,2011) MLBE:Multimodal latent binary embedding (Zhen and Yeung, 2012a) CRH:Co-regularized hashing (Zhen and Yeung,2012b) IMH:Inter-media hashing (Song et al.,2013) RaHH:Relation-aware heterogeneous hashing(Ou et al.,2013) SCM:Semantic correlation maximization (Zhang and Li,2014) CMFH:Collective matrix factorization hashing (Ding et al.,2014) QCH:Quantized correlation hashing (Wu et al.,2015) o SePH:Semantics-preserving hashing (Lin et al.,2015b) Li (http://cs.nju.edu.cn/lvj) Learning to Hash CS.NJU 18/210
Introduction Cross-Modal Hashing Given a query of either image or text, return images or texts similar to it. CVH: Cross view hashing (Kumar and Udupa, 2011) MLBE: Multimodal latent binary embedding (Zhen and Yeung, 2012a) CRH: Co-regularized hashing (Zhen and Yeung, 2012b) IMH: Inter-media hashing (Song et al., 2013) RaHH: Relation-aware heterogeneous hashing (Ou et al., 2013) SCM: Semantic correlation maximization (Zhang and Li, 2014) CMFH: Collective matrix factorization hashing (Ding et al., 2014) QCH: Quantized correlation hashing (Wu et al., 2015) SePH: Semantics-preserving hashing (Lin et al., 2015b) Li (http://cs.nju.edu.cn/lwj) Learning to Hash CS, NJU 18 / 210
Introduction Deep Hashing Deep learning for hashing o CNNH:Supervised hashing via image representation learning(Xia etal.,2014) NINH:Simultaneous feature learning and hash coding with deep neural networks(Lai et al.,2015) DSRH:Deep semantic ranking based hashing(Zhao et al.,2015) DRSCH:Bit-scalable deep hashing (Zhang et al.,2015) DH:Deep hashing for compact binary codes learning (Liong et al., 2015) Deep learning of binary hash codes (Lin et al.,2015a) o DPSH:Feature learning based deep supervised hashing with pairwise labels (Li et al.,2015) Li (http://cs.nju.edu.cn/lvj) Learning to Hash C5.NJU19/210
Introduction Deep Hashing Deep learning for hashing CNNH: Supervised hashing via image representation learning (Xia et al., 2014) NINH: Simultaneous feature learning and hash coding with deep neural networks (Lai et al., 2015) DSRH: Deep semantic ranking based hashing (Zhao et al., 2015) DRSCH: Bit-scalable deep hashing (Zhang et al., 2015) DH: Deep hashing for compact binary codes learning (Liong et al., 2015) Deep learning of binary hash codes (Lin et al., 2015a) DPSH: Feature learning based deep supervised hashing with pairwise labels (Li et al., 2015) Li (http://cs.nju.edu.cn/lwj) Learning to Hash CS, NJU 19 / 210
Introduction Quantization The quantization stage is at least as important as the projection stage: DBQ:Double-bit quantization(Kong and Li,2012a) MQ:Manhattan quantization (Kong et al.,2012) VBQ:Variable bit quantization (Moran et al.,2013) 日卡024元，互Q0 Li (http://cs.nju.edu.cn/lwj) Learning to Hash CS.NJU 20/210
Introduction Quantization The quantization stage is at least as important as the projection stage: DBQ: Double-bit quantization (Kong and Li, 2012a) MQ: Manhattan quantization (Kong et al., 2012) VBQ: Variable bit quantization (Moran et al., 2013) Li (http://cs.nju.edu.cn/lwj) Learning to Hash CS, NJU 20 / 210