3rd Place Solution for VisDA 2021 Challenge -- Universally Domain Adaptive Image Recognition
This work addresses domain adaptation challenges in computer vision, but it is incremental as it aggregates existing methods.
The paper tackled the problem of universal domain adaptation for image recognition, addressing both distribution shift and label set variance, and achieved 3rd place in the VisDA 2021 Challenge with 48.49% accuracy and 70.8% AUROC.
The Visual Domain Adaptation (VisDA) 2021 Challenge calls for unsupervised domain adaptation (UDA) methods that can deal with both input distribution shift and label set variance between the source and target domains. In this report, we introduce a universal domain adaptation (UniDA) method by aggregating several popular feature extraction and domain adaptation schemes. First, we utilize VOLO, a Transformer-based architecture with state-of-the-art performance in several visual tasks, as the backbone to extract effective feature representations. Second, we modify the open-set classifier of OVANet to recognize the unknown class with competitive accuracy and robustness. As shown in the leaderboard, our proposed UniDA method ranks the 3rd place with 48.49% ACC and 70.8% AUROC in the VisDA 2021 Challenge.