Cross-View Feature Selection

Exploring features for complicated objects: cross-view feature selection for multi-instance learning
In traditional multi-instance learning (MIL), instances are typically represented by using a single feature view. As MIL becoming …
DOI