Timezone: »
In visual domain adaptation (DA), separating the domain-specific characteristics from the domain-invariant representations is an ill-posed problem. Existing methods apply different kinds of priors or directly minimize the domain discrepancy to address this problem, which lack flexibility in handling real-world situations. Another research pipeline expresses the domain-specific information as a gradual transferring process, which tends to be suboptimal in accurately removing the domain-specific properties. In this paper, we address the modeling of domain-invariant and domain-specific information from the heuristic search perspective. We identify the characteristics in the existing representations that lead to larger domain discrepancy as the heuristic representations. With the guidance of heuristic representations, we formulate a principled framework of Heuristic Domain Adaptation (HDA) with well-founded theoretical guarantees. To perform HDA, the cosine similarity scores and independence measurements between domain-invariant and domain-specific representations are cast into the constraints at the initial and final states during the learning procedure. Similar to the final condition of heuristic search, we further derive a constraint enforcing the final range of heuristic network output to be small. Accordingly, we propose Heuristic Domain Adaptation Network (HDAN), which explicitly learns the domain-invariant and domain-specific representations with the above mentioned constraints. Extensive experiments show that HDAN has exceeded state-of-the-art on unsupervised DA, multi-source DA and semi-supervised DA. The code is available at https://github.com/cuishuhao/HDA.
Author Information
Shuhao Cui (ICT UCAS)
Xuan Jin (Alibaba Turing Lab, Alibaba Group)
Shuhui Wang (VIPL,ICT,Chinese academic of science)
Yuan He (Alibaba Group)
Qingming Huang (University of Chinese Academy of Sciences)
More from the Same Authors
-
2019 Poster: Generalized Block-Diagonal Structure Pursuit: Learning Soft Latent Task Assignment against Negative Transfer »
Zhiyong Yang · Qianqian Xu · Yangbangyan Jiang · Xiaochun Cao · Qingming Huang -
2019 Poster: DM2C: Deep Mixed-Modal Clustering »
Yangbangyan Jiang · Qianqian Xu · Zhiyong Yang · Xiaochun Cao · Qingming Huang -
2019 Spotlight: DM2C: Deep Mixed-Modal Clustering »
Yangbangyan Jiang · Qianqian Xu · Zhiyong Yang · Xiaochun Cao · Qingming Huang -
2019 Poster: iSplit LBI: Individualized Partial Ranking with Ties via Split LBI »
Qianqian Xu · Xinwei Sun · Zhiyong Yang · Xiaochun Cao · Qingming Huang · Yuan Yao