• Home
  • Search
  • Linear Discriminant Analysis via Pseudo Labels: A Unified Framework for Visual Domain Adaptation
  • Cite Icon18
  • https://doi.org/10.1109/access.2020.3035422Copy DOI Icon

Linear Discriminant Analysis via Pseudo Labels: A Unified Framework for Visual Domain Adaptation

Show More
  • Abstract
  • Highlights & Summary
  • PDF
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

This paper deals with the problem of visual domain adaptation in which source domain labeled data is available for training, but the target domain unlabeled data is available for testing. Many recent domain adaptation methods merely concentrate on extracting domain-invariant features via minimizing the distributional and geometrical divergence between domains simultaneously while ignoring within-class and between-class structure properties, especially for the target domain due to the unavailability of labeled data. We propose Linear Discriminant Analysis via Pseudo Labels (LDAPL), a unified framework for visual domain adaptation that can tackle these two issues together. LDAPL is to learn a domain-invariant features across both domains with preserving important properties such as minimizing the shift between domains both statistically and geometrically, retaining the original similarity of data samples, maximizing the target domain variance, and minimizing the within-class and maximizing the between-class properties of both domains. Specifically, LDAPL preserves the target domain discriminative information (or within-class and between-class properties) using pseudo labels, and these pseudo labels are refined until convergence. Extensive experiment on several visual cross-domain benchmarks, including Office+Caltech10 with all three types of features (such as Speeded Up Robust Features (SURF), Deep Convolutional Activation Feature (DeCAF <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">6</sub> ), and Visual Geometry Group-Fully Connected layer ( VGG-FC <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">6</sub> ) features), COIL20 (Columbia Object Image Library), digit, and PIE (Pose, Illumination, and Expression), LDAPL achieved average accuracies of 79.11%, 99.72%, 79.0%, and 84.50%, respectively. Comparative results on several visual cross-domain classification tasks verify that LDAPL can significantly outperform the state-of-the-art primitive and domain adaptation methods. Specifically, LDAPL gains over baseline Joint Geometrical and Statistical Alignment (JGSA) method with 6.6%, 5.3%, 6.3%, and 44.93% average accuracies on Office+Caltech10 (SURF, DeCAF <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">6</sub> , and VGG-FC6 ), COIL20, digit, and PIE, respectively.

Loading PDF

Similar Papers
  • Conference Article
  • Citations5

Statistical and Geometrical Alignment using Metric Learning in Domain Adaptation

  • Jul 01, 2020
  • Rakesh Kumar Sanodiya +3
  • Research Article
  • Citations8

A unified framework for visual domain adaptation with covariance matching

  • Aug 12, 2023
  • Knowledge-Based Systems
  • Ravi Ranjan Prasad Karn +2
  • Research Article
  • Citations29

Supervised domain adaptation of decision forests: Transfer of models trained in vitro for in vivo intravascular ultrasound tissue characterization.

  • Mar 17, 2016
  • Medical Image Analysis
  • Sailesh Conjeti +7
  • Research Article
  • Citations17

Manifold embedded joint geometrical and statistical alignment for visual domain adaptation

  • Sep 22, 2022
  • Knowledge-Based Systems
  • Rakesh Kumar Sanodiya +3
  • Conference Article
  • Citations4

A Feature and Parameter Selection Approach for Visual Domain Adaptation using Particle Swarm Optimization

  • Jul 18, 2022
  • Ravi Ranjan Prasad Karn +6
  • Book Chapter
  • Citations2

Unified Framework for Visual Domain Adaptation Using Globality-Locality Preserving Projections

  • Jan 01, 2019
  • Rakesh Kumar Sanodiya +2
  • Research Article
  • Citations80

Class-Aligned and Class-Balancing Generative Domain Adaptation for Hyperspectral Image Classification

  • Jan 01, 2024
  • IEEE Transactions on Geoscience and Remote Sensing
  • Jie Feng +6
  • Research Article
  • Citations42

Improving pseudo labels with intra-class similarity for unsupervised domain adaptation

  • Feb 02, 2023
  • Pattern Recognition
  • Jie Wang +1
  • Research Article
  • Citations43

Pseudo Labels for Unsupervised Domain Adaptation: A Review

  • Aug 03, 2023
  • Electronics
  • Yundong Li +2
  • Research Article
  • Citations37

Source data‐free domain adaptation of object detector through domain‐specific perturbation

  • May 08, 2021
  • International Journal of Intelligent Systems
  • Lin Xiong +5
  • Conference Article
  • Citations1

Source-free Domain Adaptation via Multicentric Prototype for Alzheimer's Disease Detection

  • Jan 06, 2023
  • Qiongmin Zhang +2
  • Book Chapter

Iterative Discriminative Domain Adaptation

  • Jan 01, 2019
  • Xiaofu Wu +3
  • Research Article
  • Citations6

A two-stage multi-scale domain adversarial transfer learning method and its application in fault diagnosis

  • Sep 18, 2023
  • Measurement Science and Technology
  • Mingyuan Zhang +5
  • Book Chapter
  • Citations7

A Feature Selection Approach to Visual Domain Adaptation in Classification

  • Jan 01, 2020
  • Rakesh Kumar Sanodiya +4
  • Research Article
  • Citations32

Feature concatenation for adversarial domain adaptation

  • Dec 13, 2020
  • Expert Systems with Applications
  • Jingyao Li +2
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.