Learning Semantic Representations for Unsupervised Domain Adaptation

By Shaoan Xie, Zibin Zheng, Liang Chen, Chuan Chen. 2018

535 citations recorded by the source corpus.

Subfields: Representation & Self-Supervised Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Representation & Self-Supervised Learning

1. Jian Liang
2. Mingsheng Long
3. Zhangjie Cao
4. Jianmin Wang
5. Jiashi Feng
6. Dacheng Tao
7. Dapeng Hu
8. K. Jia

Citations in the corpus

• Learning Robust Global Representations by Penalizing Local Predictive Power (2019, background): “…and coercing domain invariance via adversarial learning (or further extensions) (Ganin et al., 2016; Bousmalis et al., 2017; Tzeng et al., 2017; Xie et al., 2018; Hoffman et al., 2018; Long et al., 2018; Zhao et al., 2018b; Kumar et al., 2018; Li et al., 2018b; Zhao et al., 2018a;…”
• Continual Semantic Segmentation via Repulsion-Attraction of Sparse and Disentangled Latent Representations (2021, methodology): “Prototypes-based regularizing terms gained a great interest and, in particular, have been largely used in the literature of few-shot learning [15, 63, 60], to learn prototypical representations of each category, and domain adaptation, to enforce orthogonality [50, 65] or centroid matching [70, 12].”
• Universal Domain Adaptation (2019, background): “deep adaptation methods further explore architecture designs [19, 43, 36, 26, 20, 41, 16, 47, 25, 4, 18].”
• GANcraft: Unsupervised 3D Neural Rendering of Minecraft Worlds (2021, background): “3 [65] Shaoan Xie, Zibin Zheng, Liang Chen, and Chuan Chen.”
• Transfer Learning with Dynamic Adversarial Adaptation Network (2019, background): “Recently, we have witnessed considerable research [18], [25], [37], [38] for adversarial domain adaptation.”
• Domain Adaptation for Semantic Segmentation With Maximum Squares Loss (2019, background): “Traditionally, un-supervised domain adaptation (UDA) [10, 20, 21, 29, 30, 33] is studied to tackle the domain-shift problem between the labeled source domain and unlabeled target domain for the classification task.”
• Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation (2020, result): “The intuition behind is quite similar to previous studies (Long et al., 2013; Xie et al., 2018) where a simplified MMD is exploited for discriminative domain confusion.”
• Domain-Symmetric Networks for Adversarial Domain Adaptation (2019, methodology): “In this section, we briefly review recent domain adaptation methods, in particular those aiming to align the joint distributions of feature and category across two data domains [22, 34, 30, 13, 23, 12, 18].”
• HGNet: Hybrid Generative Network for Zero-Shot Domain Adaptation (2020, background): “Concretely, it attempts to borrow knowledge from well-annotated modality (source domain) to solve classification task on target domain without any label information [30, 32, 35].”
• Adversarial Domain Adaptation with Domain Mixup (2019, background): “In addition, the label information of target domain is explored in recent works (Zhang et al. 2018b; Xie et al. 2018; Ma, Zhang, and Xu 2019).”
• Reducing Domain Gap via Style-Agnostic Networks (2019, methodology): “However, MSTN [28] and clustering-based pseudo labels [1] have not shown clear improvements in our experiments, possibly due to the high complexity of the given task.”
• Harmonizing Transferability and Discriminability for Adapting Object Detectors (2020, background): “, prototype alignment [58, 5, 37, 10] or entropy regularization [64, 47, 8]) would be no longer applicable.”
• Deep Subdomain Adaptation Network for Image Classification (2020, background): “MSTN [20] learns semantic representations for unlabeled target samples by aligning labeled source centroid and pseudo-labeled target centroid.”
• Model Adaptation: Unsupervised Domain Adaptation Without Source Data (2020, background): “The expected error on the target domain is bounded by: 1) the expected error on the source domain; 2) the domain discrepancy between the source and target domains; and 3) a shared expected loss which is expected to be small [60].”
• Reducing Domain Gap by Reducing Style Bias (2021, methodology): “To fully leverage those unlabeled data, we utilize a simple extension of SagNets based on consistency learning [37, 49].”
• Cross-Domain Gradient Discrepancy Minimization for Unsupervised Domain Adaptation (2021, methodology): “Some previous studies [43, 6, 39] directly incorporate naive pseudo labeling strategy into their methods.”
• HoMM: Higher-order Moment Matching for Unsupervised Domain Adaptation (2019, background): “Therefore, a large body of work pays attention to learn the discriminative clusters in the target domain via entropy minimization [13, 28, 34], pseudo label [21, 32, 43] or distance-based metrics [3, 18].”