Domain-Agnostic Prior for Transfer Semantic Segmentation
This work addresses domain shift in semantic segmentation for computer vision applications, offering an incremental improvement with a simple regularization mechanism.
The paper tackles unsupervised domain adaptation for semantic segmentation by introducing a domain-agnostic prior to align features across domains, achieving improved segmentation accuracy beyond state-of-the-art methods, particularly with a text embedding model.
Unsupervised domain adaptation (UDA) is an important topic in the computer vision community. The key difficulty lies in defining a common property between the source and target domains so that the source-domain features can align with the target-domain semantics. In this paper, we present a simple and effective mechanism that regularizes cross-domain representation learning with a domain-agnostic prior (DAP) that constrains the features extracted from source and target domains to align with a domain-agnostic space. In practice, this is easily implemented as an extra loss term that requires a little extra costs. In the standard evaluation protocol of transferring synthesized data to real data, we validate the effectiveness of different types of DAP, especially that borrowed from a text embedding model that shows favorable performance beyond the state-of-the-art UDA approaches in terms of segmentation accuracy. Our research reveals that UDA benefits much from better proxies, possibly from other data modalities.