Partial Video Domain Adaptation with Partial Adversarial Temporal Attentive Network
This work addresses negative transfer in video domain adaptation for scenarios with mismatched class sets, which is an incremental improvement over existing methods.
The paper tackles the problem of Partial Video Domain Adaptation (PVDA), where source and target domains share only a subset of classes, by proposing a Partial Adversarial Temporal Attentive Network (PATAN) that filters out irrelevant source classes using spatial and temporal features, achieving state-of-the-art performance on new PVDA benchmarks.
Partial Domain Adaptation (PDA) is a practical and general domain adaptation scenario, which relaxes the fully shared label space assumption such that the source label space subsumes the target one. The key challenge of PDA is the issue of negative transfer caused by source-only classes. For videos, such negative transfer could be triggered by both spatial and temporal features, which leads to a more challenging Partial Video Domain Adaptation (PVDA) problem. In this paper, we propose a novel Partial Adversarial Temporal Attentive Network (PATAN) to address the PVDA problem by utilizing both spatial and temporal features for filtering source-only classes. Besides, PATAN constructs effective overall temporal features by attending to local temporal features that contribute more toward the class filtration process. We further introduce new benchmarks to facilitate research on PVDA problems, covering a wide range of PVDA scenarios. Empirical results demonstrate the state-of-the-art performance of our proposed PATAN across the multiple PVDA benchmarks.