MetaViewer: Towards A Unified Multi-View Representation
This addresses multi-view representation learning for researchers, offering a novel approach to improve unified representations, though it appears incremental as it builds on existing multi-view methods.
The paper tackles the problem of manually pre-specified fusion functions and view-private redundant information degrading multi-view representation quality by proposing MetaViewer, a bi-level-optimization framework that learns fusion and filters out view-private information, achieving effectiveness in classification and clustering tasks.
Existing multi-view representation learning methods typically follow a specific-to-uniform pipeline, extracting latent features from each view and then fusing or aligning them to obtain the unified object representation. However, the manually pre-specify fusion functions and view-private redundant information mixed in features potentially degrade the quality of the derived representation. To overcome them, we propose a novel bi-level-optimization-based multi-view learning framework, where the representation is learned in a uniform-to-specific manner. Specifically, we train a meta-learner, namely MetaViewer, to learn fusion and model the view-shared meta representation in outer-level optimization. Start with this meta representation, view-specific base-learners are then required to rapidly reconstruct the corresponding view in inner-level. MetaViewer eventually updates by observing reconstruction processes from uniform to specific over all views, and learns an optimal fusion scheme that separates and filters out view-private information. Extensive experimental results in downstream tasks such as classification and clustering demonstrate the effectiveness of our method.