• Home
  • Search
  • SIDDA: SInkhorn Dynamic Domain Adaptation for image classification with equivariant neural networks
  • Cite Icon3
  • https://doi.org/10.1088/2632-2153/adf701Copy DOI Icon

SIDDA: SInkhorn Dynamic Domain Adaptation for image classification with equivariant neural networks

Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Abstract Modern neural networks (NNs) often do not generalize well in the presence of a "covariate shift''; that is, in situations where the training and test data distributions differ, but the conditional distribution of classification labels given the data remains unchanged. In such cases, NN generalization can be reduced to a problem of learning domain-invariant features. Domain adaptation (DA) methods include a broad range of techniques aimed at achieving this; however, these methods have struggled with the need for hyperparameter tuning, which then incurs computational costs. In this work, we introduce SIDDA, an out-of-the-box DA training algorithm built upon the Sinkhorn divergence, that can achieve effective domain alignment with minimal hyperparameter tuning and computational overhead. We demonstrate the efficacy of our method on multiple simulated and real datasets of varying complexity, including simple shapes, handwritten digits, and real astronomical observations. These datasets exhibit covariate shifts due to noise, blurring, and differences between telescopes. SIDDA is compatible with a variety of NN architectures, and it works particularly well in improving classification accuracy and model calibration when paired with equivariant NNs (ENNs). We find that SIDDA enhances the generalization capabilities of NNs, achieving up to a 40% improvement in classification accuracy on unlabeled target data. We also study the efficacy of DA on ENNs with respect to the varying orders of the dihedral group, and find that the model performance improves with higher orders. Finally, we find that SIDDA enhances model calibration on both source and target data, with the most significant gains in the unlabeled target domain---achieving over an order of magnitude improvement in the ECE and Brier score. SIDDA's versatility across various NN models and datasets, combined with its automated approach to domain alignment, has the potential to significantly advance multi-dataset studies by enabling the development of highly generalizable models

Similar Papers
  • Research Article
  • Citations45

Universal multi-Source domain adaptation for image classification

  • Aug 14, 2021
  • Pattern Recognition
  • Yueming Yin +3
  • PDF
  • Research Article
  • Citations23

Unsupervised Domain Adaptation for Image Classification and Object Detection Using Guided Transfer Learning Approach and JS Divergence.

  • Apr 30, 2023
  • Sensors
  • Parth Goel +1
  • Research Article
  • Citations2

Toward Open-World Domain Adaptation via Iteratively Contrastive Learning and Clustering.

  • Jan 01, 2025
  • IEEE transactions on neural networks and learning systems
  • Jingzheng Li +4
  • Conference Article
  • Citations31

Adversarial Robustness for Unsupervised Domain Adaptation

  • Oct 01, 2021
  • Muhammad Awais +6
  • Dissertation

Towards robust sensing and recognition : from statistical learning to transfer learning

  • Jan 01, 2020
  • Jianfei Yang
  • Dissertation
  • Citations1

Domain Adaptation and Domain Generalization with Representation Learning

  • Jan 01, 2016
  • Muhammad Ghifary
  • Dissertation
  • Citations2

Domain Adaptation and Domain Generalization with Representation Learning

  • Jan 01, 2016
  • Muhammad Ghifary
  • Dissertation

Label-efficient visual recognition with transfer learning

  • Jan 01, 2025
  • Jingyi Zhang
  • Conference Article
  • Citations3

CUDA: Contradistinguisher for Unsupervised Domain Adaptation

  • Nov 01, 2019
  • Sourabh Balgi +1
  • Research Article

Deep transferable label propagation with prototypical augmentation.

  • May 18, 2026
  • Scientific reports
  • Yufang Dan +2
  • Conference Article
  • Citations10

Partially Zero-shot Domain Adaptation from Incomplete Target Data with Missing Classes

  • Mar 01, 2020
  • Masato Ishii +2
  • Research Article
  • Citations40

Source-Free Multidomain Adaptation With Fuzzy Rule-Based Deep Neural Networks

  • Dec 01, 2023
  • IEEE Transactions on Fuzzy Systems
  • Keqiuyin Li +3
  • Research Article
  • Citations18

Leveraging topology for domain adaptive road segmentation in satellite and aerial imagery

  • Nov 09, 2023
  • ISPRS Journal of Photogrammetry and Remote Sensing
  • Javed Iqbal +3
  • Research Article
  • Citations27

Multi-Source Collaborative Contrastive Learning for Decentralized Domain Adaptation

  • May 01, 2023
  • IEEE Transactions on Circuits and Systems for Video Technology
  • Yikang Wei +3
  • Research Article
  • Citations13

Adaptive centroid prototype-based domain adaptation for fault diagnosis of rotating machinery without source data

  • Jul 31, 2024
  • Reliability Engineering and System Safety
  • Qikang Li +4
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.