COMET: Contrastive Mean Teacher for Online Source-Free Universal Domain Adaptation

• URL: http://arxiv.org/abs/2401.17728v2
• Date: Thu, 2 May 2024 06:49:35 GMT
• Title: COMET: Contrastive Mean Teacher for Online Source-Free Universal Domain Adaptation
• Authors: Pascal Schlachter, Bin Yang,
• Abstract summary: In real-world applications, there is often a domain shift from training to test data. We introduce a Contrastive Mean Teacher (COMET) tailored to this novel scenario. COMET yields state-of-the-art performance and proves to be consistent and robust across a variety of different scenarios.
• Score: 3.5139431332194198
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In real-world applications, there is often a domain shift from training to test data. This observation resulted in the development of test-time adaptation (TTA). It aims to adapt a pre-trained source model to the test data without requiring access to the source data. Thereby, most existing works are limited to the closed-set assumption, i.e. there is no category shift between source and target domain. We argue that in a realistic open-world setting a category shift can appear in addition to a domain shift. This means, individual source classes may not appear in the target domain anymore, samples of new classes may be part of the target domain or even both at the same time. Moreover, in many real-world scenarios the test data is not accessible all at once but arrives sequentially as a stream of batches demanding an immediate prediction. Hence, TTA must be applied in an online manner. To the best of our knowledge, the combination of these aspects, i.e. online source-free universal domain adaptation (online SF-UniDA), has not been studied yet. In this paper, we introduce a Contrastive Mean Teacher (COMET) tailored to this novel scenario. It applies a contrastive loss to rebuild a feature space where the samples of known classes build distinct clusters and the samples of new classes separate well from them. It is complemented by an entropy loss which ensures that the classifier output has a small entropy for samples of known classes and a large entropy for samples of new classes to be easily detected and rejected as unknown. To provide the losses with reliable pseudo labels, they are embedded into a mean teacher (MT) framework. We evaluate our method across two datasets and all category shifts to set an initial benchmark for online SF-UniDA. Thereby, COMET yields state-of-the-art performance and proves to be consistent and robust across a variety of different scenarios.

Related papers

• Memory-Efficient Pseudo-Labeling for Online Source-Free Universal Domain Adaptation using a Gaussian Mixture Model [3.1265626879839923]
  Universal domain adaptation (UniDA) has gained attention for addressing the possibility of an additional category (label) shift between the source and target domain. We propose a novel method that continuously captures the distribution of known classes in the feature space using a Gaussian mixture model (GMM) Our approach not only achieves state-of-the-art results in all experiments on the DomainNet dataset but also significantly outperforms the existing methods on the challenging VisDA-C dataset.
  arXiv  Detail & Related papers  (2024-07-19T11:13:31Z)
• Adaptive Test-Time Personalization for Federated Learning [51.25437606915392]
  We introduce a novel setting called test-time personalized federated learning (TTPFL) In TTPFL, clients locally adapt a global model in an unsupervised way without relying on any labeled data during test-time. We propose a novel algorithm called ATP to adaptively learn the adaptation rates for each module in the model from distribution shifts among source domains.
  arXiv  Detail & Related papers  (2023-10-28T20:42:47Z)
• Upcycling Models under Domain and Category Shift [95.22147885947732]
  We introduce an innovative global and local clustering learning technique (GLC) We design a novel, adaptive one-vs-all global clustering algorithm to achieve the distinction across different target classes. Remarkably, in the most challenging open-partial-set DA scenario, GLC outperforms UMAD by 14.8% on the VisDA benchmark.
  arXiv  Detail & Related papers  (2023-03-13T13:44:04Z)
• CAFA: Class-Aware Feature Alignment for Test-Time Adaptation [50.26963784271912]
  Test-time adaptation (TTA) aims to address this challenge by adapting a model to unlabeled data at test time. We propose a simple yet effective feature alignment loss, termed as Class-Aware Feature Alignment (CAFA), which simultaneously encourages a model to learn target representations in a class-discriminative manner.
  arXiv  Detail & Related papers  (2022-06-01T03:02:07Z)
• On Universal Black-Box Domain Adaptation [53.7611757926922]
  We study an arguably least restrictive setting of domain adaptation in a sense of practical deployment. Only the interface of source model is available to the target domain, and where the label-space relations between the two domains are allowed to be different and unknown. We propose to unify them into a self-training framework, regularized by consistency of predictions in local neighborhoods of target samples.
  arXiv  Detail & Related papers  (2021-04-10T02:21:09Z)
• OVANet: One-vs-All Network for Universal Domain Adaptation [78.86047802107025]
  Existing methods manually set a threshold to reject unknown samples based on validation or a pre-defined ratio of unknown samples. We propose a method to learn the threshold using source samples and to adapt it to the target domain. Our idea is that a minimum inter-class distance in the source domain should be a good threshold to decide between known or unknown in the target.
  arXiv  Detail & Related papers  (2021-04-07T18:36:31Z)
• Domain Impression: A Source Data Free Domain Adaptation Method [27.19677042654432]
  Unsupervised domain adaptation methods solve the adaptation problem for an unlabeled target set, assuming that the source dataset is available with all labels. This paper proposes a domain adaptation technique that does not need any source data. Instead of the source data, we are only provided with a classifier that is trained on the source data.
  arXiv  Detail & Related papers  (2021-02-17T19:50:49Z)
• A Balanced and Uncertainty-aware Approach for Partial Domain Adaptation [142.31610972922067]
  This work addresses the unsupervised domain adaptation problem, especially in the case of class labels in the target domain being only a subset of those in the source domain. We build on domain adversarial learning and propose a novel domain adaptation method BA$3$US with two new techniques termed Balanced Adversarial Alignment (BAA) and Adaptive Uncertainty Suppression (AUS) Experimental results on multiple benchmarks demonstrate our BA$3$US surpasses state-of-the-arts for partial domain adaptation tasks.
  arXiv  Detail & Related papers  (2020-03-05T11:37:06Z)
• Enlarging Discriminative Power by Adding an Extra Class in Unsupervised Domain Adaptation [5.377369521932011]
  We propose an idea of empowering the discriminativeness: Adding a new, artificial class and training the model on the data together with the GAN-generated samples of the new class. Our idea is highly generic so that it is compatible with many existing methods such as DANN, VADA, and DIRT-T.
  arXiv  Detail & Related papers  (2020-02-19T07:58:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.