Article Review Deep Correspondence Restricted Boltzmann...

Article Review : Deep correspondence restricted Boltzmann machine for cross-modal retrieval Review Submission : ACN 5314.5H1 - Computational Modeling Methods in Behavioral Brain Sci. Reviewer : Jithin Pradeep R jxp161430@utdallas.edu School of Behavioral and Brain Science, The University of Texas at Dallas December 16, 2016. Deep correspondence restricted Boltzmann machine for cross-modal retrieval: Jithin Pradeep Article Review. Article Review : Deep correspondence restricted Boltzmann machine for cross-modal retrieval Abstract of article Cross-modal retrieval task tries to exploit the correlation between the component using a canonical cor-relational analysis. In simple word, cross model retrieval would involve retrieving an image using a text input or image to generate a corresponding narration. In world where internet user throws up bunch of multimodal content make it important to analysis the same. Modeling the correlations between dierent modalities is the key to tackle cross model retrieval problem. In the paper,author propose a correspondence restricted Boltzmann machine(Corr-RBM) to map the original features of bi-modal data, such as image and text, into a low-dimensional common space by deploying two deep neural structures using Corr-RBM as the main building block for the task of cross-modal retrieval. The heterogeneous data are made comparable by optimizing a single objective function (constructed to trade o the correlation loss and likelihoods of both