Partial Domain Adaptation Using Selective Representation Learning For Class-Weight Computation
This work tackles the practical challenge of partial domain adaptation, which is relevant for researchers and practitioners dealing with limited labeled data in real-world scenarios.
This paper addresses partial domain adaptation where the source label set is a superset of the target label set. The authors propose a method that identifies outlier classes based on image content and trains a label classifier on class-content from source images, transforming soft class-level weights into two clusters to eliminate negative transfer.
The generalization power of deep-learning models is dependent on rich-labelled data. This supervision using large-scaled annotated information is restrictive in most real-world scenarios where data collection and their annotation involve huge cost. Various domain adaptation techniques exist in literature that bridge this distribution discrepancy. However, a majority of these models require the label sets of both the domains to be identical. To tackle a more practical and challenging scenario, we formulate the problem statement from a partial domain adaptation perspective, where the source label set is a super set of the target label set. Driven by the motivation that image styles are private to each domain, in this work, we develop a method that identifies outlier classes exclusively from image content information and train a label classifier exclusively on class-content from source images. Additionally, elimination of negative transfer of samples from classes private to the source domain is achieved by transforming the soft class-level weights into two clusters, 0 (outlier source classes) and 1 (shared classes) by maximizing the between-cluster variance between them.