TR2022-134

Incomplete Multi-view Domain Adaptation via Channel Enhancement and Knowledge Transfer


    •  Xia, H., Wang, P., Ding, Z., "Incomplete Multi-view Domain Adaptation via Channel Enhancement and Knowledge Transfer", European Conference on Computer Vision (ECCV), October 2022.
      BibTeX TR2022-134 PDF
      • @inproceedings{Xia2022oct,
      • author = {Xia, Haifeng and Wang, Pu and Ding, Zhengming},
      • title = {Incomplete Multi-view Domain Adaptation via Channel Enhancement and Knowledge Transfer},
      • booktitle = {European Conference on Computer Vision (ECCV)},
      • year = 2022,
      • month = oct,
      • url = {https://www.merl.com/publications/TR2022-134}
      • }
  • MERL Contact:
  • Research Areas:

    Artificial Intelligence, Communications, Machine Learning, Optimization, Signal Processing

Abstract:

Unsupervised domain adaptation (UDA) borrows well-labeled source knowledge to solve the specific task on unlabeled target domain with the assumption that both domains are from a single sensor, e.g., RGB or depth images. To boost model performance, multiple sensors are deployed on new-produced devices like autonomous vehicles to benefit from enriched information. However, the model trained with multi- view data difficultly becomes compatible with conventional devices only with a single sensor. This scenario is defined as incomplete multi-view domain adaptation (IMVDA), which considers that the source domain consists of multi-view data while the target domain only includes single- view instances. To overcome this practical demand, this paper proposes a novel Channel Enhancement and Knowledge Transfer (CEKT) frame- work with two modules. Concretely, the source channel enhancement module distinguishes view-common from view-specific channels and explores channel similarity to magnify the representation of important channels. Moreover, the adaptive knowledge transfer module attempts to enhance target representation towards multi-view semantic through implicit missing view recovery and adaptive cross-domain alignment. Extensive experimental results illustrate the effectiveness of our method in solving the IMVDA challenge.