Online Continual Adaptation with Active Self-Training

• URL: http://arxiv.org/abs/2106.06526v1
• Date: Fri, 11 Jun 2021 17:51:25 GMT
• Title: Online Continual Adaptation with Active Self-Training
• Authors: Shiji Zhou, Han Zhao, Shanghang Zhang, Lianzhe Wang, Heng Chang, Zhi Wang, Wenwu Zhu
• Abstract summary: We propose an online setting where the learner aims to continually adapt to changing distributions using both unlabeled samples and active queries of limited labels. Online Self-Adaptive Mirror Descent (OSAMD) adopts an online teacher-student structure to enable online self-training from unlabeled data. We show that OSAMD achieves favorable regrets under changing environments with limited labels on both simulated and real-world data.
• Score: 69.5815645379945
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Models trained with offline data often suffer from continual distribution shifts and expensive labeling in changing environments. This calls for a new online learning paradigm where the learner can continually adapt to changing environments with limited labels. In this paper, we propose a new online setting -- Online Active Continual Adaptation, where the learner aims to continually adapt to changing distributions using both unlabeled samples and active queries of limited labels. To this end, we propose Online Self-Adaptive Mirror Descent (OSAMD), which adopts an online teacher-student structure to enable online self-training from unlabeled data, and a margin-based criterion that decides whether to query the labels to track changing distributions. Theoretically, we show that, in the separable case, OSAMD has an $O({T}^{1/2})$ dynamic regret bound under mild assumptions, which is even tighter than the lower bound $\Omega(T^{2/3})$ of traditional online learning with full labels. In the general case, we show a regret bound of $O({\alpha^*}^{1/3} {T}^{2/3} + \alpha^* T)$, where $\alpha^*$ denotes the separability of domains and is usually small. Our theoretical results show that OSAMD can fast adapt to changing environments with active queries. Empirically, we demonstrate that OSAMD achieves favorable regrets under changing environments with limited labels on both simulated and real-world data, which corroborates our theoretical findings.

Related papers

• COMET: Contrastive Mean Teacher for Online Source-Free Universal Domain Adaptation [3.5139431332194198]
  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.
  arXiv  Detail & Related papers  (2024-01-31T10:47:25Z)
• Towards Real-World Test-Time Adaptation: Tri-Net Self-Training with Balanced Normalization [52.03927261909813]
  Existing works mainly consider real-world test-time adaptation under non-i.i.d. data stream and continual domain shift. We argue failure of state-of-the-art methods is first caused by indiscriminately adapting normalization layers to imbalanced testing data. The final TTA model, termed as TRIBE, is built upon a tri-net architecture with balanced batchnorm layers.
  arXiv  Detail & Related papers  (2023-09-26T14:06:26Z)
• Online Label Shift: Optimal Dynamic Regret meets Practical Algorithms [33.61487362513345]
  This paper focuses on supervised and unsupervised online label shift, where the class marginals $Q(y)$ varies but the class-conditionals $Q(x|y)$ remain invariant. In the unsupervised setting, our goal is to adapt a learner, trained on some offline labeled data, to changing label distributions given unlabeled online data. We develop novel algorithms that reduce the adaptation problem to online regression and guarantee optimal dynamic regret without any prior knowledge of the extent of drift in the label distribution.
  arXiv  Detail & Related papers  (2023-05-31T05:39:52Z)
• Label-Retrieval-Augmented Diffusion Models for Learning from Noisy Labels [61.97359362447732]
  Learning from noisy labels is an important and long-standing problem in machine learning for real applications. In this paper, we reformulate the label-noise problem from a generative-model perspective. Our model achieves new state-of-the-art (SOTA) results on all the standard real-world benchmark datasets.
  arXiv  Detail & Related papers  (2023-05-31T03:01:36Z)
• Adapting to Online Label Shift with Provable Guarantees [137.89382409682233]
  We formulate and investigate the problem of online label shift. The non-stationarity and lack of supervision make the problem challenging to be tackled. Our algorithms enjoy optimal dynamic regret, indicating that performance is competitive with a clairvoyant nature.
  arXiv  Detail & Related papers  (2022-07-05T15:43:14Z)
• Active learning for online training in imbalanced data streams under cold start [0.8155575318208631]
  We propose an Active Learning (AL) annotation system for datasets with orders of magnitude of class imbalance. We present a computationally efficient Outlier-based Discriminative AL approach (ODAL) and design a novel 3-stage sequence of AL labeling policies. The results show that our method can more quickly reach a high performance model than standard AL policies.
  arXiv  Detail & Related papers  (2021-07-16T06:49:20Z)
• Online Adaptation to Label Distribution Shift [37.91472909652585]
  We propose adaptation algorithms inspired by classical online learning techniques such as Follow The Leader (FTL) and Online Gradient Descent (OGD) We empirically verify our findings under both simulated and real world label distribution shifts and show that OGD is particularly effective and robust to a variety of challenging label shift scenarios.
  arXiv  Detail & Related papers  (2021-07-09T16:12:19Z)
• Group-aware Label Transfer for Domain Adaptive Person Re-identification [179.816105255584]
  Unsupervised Adaptive Domain (UDA) person re-identification (ReID) aims at adapting the model trained on a labeled source-domain dataset to a target-domain dataset without any further annotations. Most successful UDA-ReID approaches combine clustering-based pseudo-label prediction with representation learning and perform the two steps in an alternating fashion. We propose a Group-aware Label Transfer (GLT) algorithm, which enables the online interaction and mutual promotion of pseudo-label prediction and representation learning.
  arXiv  Detail & Related papers  (2021-03-23T07:57:39Z)
• Active Online Learning with Hidden Shifting Domains [64.75186088512034]
  We propose a surprisingly simple algorithm that adaptively balances its regret and its number of label queries. Our algorithm can adaptively deal with interleaving spans of inputs from different domains.
  arXiv  Detail & Related papers  (2020-06-25T15:23:59Z)

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.