Progressive Feature Alignment for Unsupervised Domain Adaptation

By Chaoqi Chen, Weiping Xie, Tingyang Xu, Wenbing Huang, Yu Rong, Xinghao Ding, Yue Huang, Junzhou Huang. 2018

481 citations recorded by the source corpus.

Unsupervised domain adaptation (UDA) transfers knowledge from a label-rich source domain to a fully-unlabeled target domain. To tackle this task, recent approaches resort to discriminative domain transfer in virtue of pseudo-labels to enforce the class-level distribution alignment across the source and target domains. These methods, however, are vulnerable to the error accumulation and thus incapable of preserving cross-domain category consistency, as the pseudo-labeling accuracy is not guaranteed explicitly. In this paper, we propose the Progressive Feature Alignment Network (PFAN) to align the discriminative features across domains progressively and effectively, via exploiting the intra-class variation in the target domain. To be specific, we first develop an Easy-to-Hard Transfer Strategy (EHTS) and an Adaptive Prototype Alignment (APA) step to train our model iteratively and alternatively. Moreover, upon observing that a good domain adaptation usually requires a non-saturated source classifier, we consider a simple yet efficient way to retard the convergence speed of the source classification loss by further involving a temperature variate into the soft-max function. The extensive experimental results reveal that the proposed PFAN exceeds the state-of-the-art performance on three UDA datasets.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Unsupervised Model Adaptation for Continual Semantic Segmentation (2020, methodology): “We have compared our performance against the adversarial learning-based UDA methods: GIO-Ada (Chen et al. 2019b), ADVENT (Vu et al. 2019), AdaSegNet (Tsai et al. 2018), TGCF-DA+SE (Choi, Kim, and Kim 2019), PCEDA (Yang et al. 2020), and CyCADA (Hoffman et al. 2018a).”
Unsupervised Domain Adaptation via Structurally Regularized Deep Clustering (2020, background): “A simple strategy to implement this hypothesis is to re-weight source instances based on their similarities to target ones [7, 17, 67].”
Racial Faces in the Wild: Reducing Racial Bias by Information Maximization Adaptation Network (2018, methodology): “Previous UDA methods apply the source classiﬁer to predict pseudo-labels in the target domain, by which the network can be ﬁne-tuned using supervised losses [51, 69, 16, 18, 66].”
Sequential Unsupervised Domain Adaptation through Prototypical Distributions (2020, background): “The mean for each mode can be considered as the class prototype in the previous works [26, 27].”
Category Anchor-Guided Unsupervised Domain Adaptation for Semantic Segmentation (2019, methodology): “Previous methods have tried to learn domain-invariant representations by matching the distributions between source and target domains at the appearance level [27, 34, 40, 13, 21], feature level [14, 27, 3, 13], or output level [45, 36, 26].”
Domain Consensus Clustering for Universal Domain Adaptation (2021, background): “Some recent works take advantage of adversarial training to promote the alignment in the input space [11, 23, 25, 20, 10] or feature space [58, 44, 6, 18, 38, 64, 43].”
Harmonizing Transferability and Discriminability for Adapting Object Detectors (2020, methodology): “A general practice in UDA is to bridge the domain gap by explicitly learning invariant representations between domains and achieving small error on the source domain, which have achieved compelling performance on image classification [15, 55, 13, 49, 54, 45, 23, 5] and semantic segmentation [52, 19,”
Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation (2021, background): “However, aligning global distribution cannot guarantee a small expected error on the target domain [7, 28, 72].”
Cross-Domain Similarity Learning for Face Recognition in Unseen Domains (2021, background): “In the future, we will investigate how a universal form of covariance matrix (such as the one used in [5]) could be utilized in our framework.”
Undoing the Damage of Label Shift for Cross-domain Semantic Segmentation (2022, background): “Recently, researchers have found that this alignment strategy may lead to the phenomenon of misalignment [7, 24].”
Instance Adaptive Self-Training for Unsupervised Domain Adaptation (2020, background): “More recently, an alternative research line to reduce domain shift focuses on building schemes based on the self-training (ST) framework [36,2,34,35,20,31].”
Confidence Regularized Self-Training (2019, background): “More recently, self-training with networks emerged as a promising alternative towards domain adaptation [4, 5, 25, 29, 49, 54, 69].”
Perpetual Humanoid Control for Real-time Simulated Avatars (2023, background): “A paradigm has been studied in transfer learning and domain adaption as progressive learning [7, 5] or curriculum learning [1].”
Classes Matter: A Fine-grained Adversarial Approach to Cross-domain Semantic Segmentation (2020, background): “o address the performance drop caused by the different distributions of training data and testing data. In the recent years, several works are proposed to approach this problem in image classication [3,25]. Inspired by the theoretical upper bound of risk in target domain [2], some pioneering works s”
Continual Test-Time Domain Adaptation (2022, background): “Therefore, early prediction mistakes are more likely to result in error accumulation [4].”