Goto

Collaborating Authors

 Asia








CPM-Nets: CrossPartialMulti-ViewNetworks

Neural Information Processing Systems

Several methods are proposed to keep on exploiting the correlation of different views. One straightforward way is completing the missing views,andthentheon-shelf multi-viewlearning algorithmscould beadopted. Themissing views are basically blockwise and thus low-rank based completion [12, 13] is not applicable which has been widely recognized [5, 14]. Missing modality imputation methods [15, 5] usually require samples with two paired modalities to train the networks which can predict the missing modality fromtheobservedone.




TightMutualInformationEstimationWith ContrastiveFenchel-LegendreOptimization

Neural Information Processing Systems

Successful applications ofInfoNCE (Information Noise-ContrastiveEstimation) and its variants have popularized the use of contrastive variational mutual information (MI) estimators in machine learning.