Source-Guided Similarity Preservation for Online Person Re-Identification
This work addresses incremental adaptation in person re-identification for surveillance applications, but it is incremental as it builds on existing UDA methods.
The paper tackles the challenges of catastrophic forgetting and domain shift in online unsupervised domain adaptation for person re-identification by proposing a Source-guided Similarity Preservation (S2P) framework, which outperforms previous state-of-the-art methods on multiple benchmarks.
Online Unsupervised Domain Adaptation (OUDA) for person Re-Identification (Re-ID) is the task of continuously adapting a model trained on a well-annotated source domain dataset to a target domain observed as a data stream. In OUDA, person Re-ID models face two main challenges: catastrophic forgetting and domain shift. In this work, we propose a new Source-guided Similarity Preservation (S2P) framework to alleviate these two problems. Our framework is based on the extraction of a support set composed of source images that maximizes the similarity with the target data. This support set is used to identify feature similarities that must be preserved during the learning process. S2P can incorporate multiple existing UDA methods to mitigate catastrophic forgetting. Our experiments show that S2P outperforms previous state-of-the-art methods on multiple real-to-real and synthetic-to-real challenging OUDA benchmarks.