• Home
  • Search
  • PEAMATL: A Strategy for Developing Near-Infrared Spectral Prediction Models Under Domain Shift Using Self-Supervised Transfer Learning
  • Cite Icon11
  • https://doi.org/10.1109/tim.2023.3273689Copy DOI Icon

PEAMATL: A Strategy for Developing Near-Infrared Spectral Prediction Models Under Domain Shift Using Self-Supervised Transfer Learning

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

Near-infrared (NIR) spectroscopy combined with spectra prediction models have been widely employed as quick and cost-effective analytical techniques in the pharmaceutical, chemical, and food industries. However, calibration has to be conducted for a prediction model being constructed on data from the source domain if we want to apply the model to a new target domain. Most deep transfer learning methods, which extract domain-invariant features from the source domain samples and transfer these features to enhance the representation ability for target domain data, are available to calibrate prediction models. However, due to the difficulty of measuring samples’ reference values (label), the reliance on labeled samples for supervised techniques to extract domain-invariant features remains a major bottleneck. In this study, we propose a novel self-supervised transfer learning approach named PEAMATL for learning and transferring generalized domain-invariant features from samples’ spectra, aiming to accurately predict unseen samples’ reference values. PEAMATL firstly trains a pyramid encoder consisting of three external attention modules to extract multi-scale features from unlabeled source domain samples using a self-supervised learning framework; secondly, it transfers the pre-trained spectra encoder followed by an initialized prediction head network to build a prediction model; finally, PEAMATL refines the model parameters using a portion of the labeled target domain samples to adapt to unseen target domain samples. The calibration analysis is tested on tablet, melamine, and apple datasets for predicting active pharmaceutical ingredient, turbidity point, soluble solid content, and firmness. Compared with three existing supervised and two self-supervised transfer learning methods, the proposed PEAMATL method achieves at least 3.32 % ~ 30.88 % prediction error reduction on 19 out of 20 scenarios involving three types of domain shift. Therefore, PEAMATL has the potential to be a generic framework for tackling the common problem of domain shift-induced performance degradation of prediction models in the domain of NIR-based quantitative analysis.

Similar Papers
  • Research Article
  • Citations25

Discriminative Fisher Embedding Dictionary Transfer Learning for Object Recognition.

  • Jan 01, 2023
  • IEEE Transactions on Neural Networks and Learning Systems
  • Zizhu Fan +4
  • Conference Article
  • Citations22

DASA: Domain adaptation in stacked autoencoders using systematic dropout

  • Nov 01, 2015
  • Abhijit Guha Roy +1
  • Research Article
  • Citations8

Alignment subdomain-based deep convolutional transfer learning for machinery fault diagnosis under different working conditions

  • Feb 04, 2022
  • Measurement Science and Technology
  • Yibing Li +2
  • Research Article
  • Citations4

Enhancing Cross-Domain Remote Sensing Scene Classification by Multi-Source Subdomain Distribution Alignment Network

  • Apr 05, 2025
  • Remote Sensing
  • Yong Wang +6
  • PDF
  • Research Article

Learning Transferable Convolutional Proxy by SMI-Based Matching Technique

  • Oct 14, 2020
  • Shock and Vibration
  • Wei Jin +1
  • Research Article
  • Citations127

Bearing fault diagnosis using transfer learning and optimized deep belief network

  • Mar 07, 2022
  • Measurement Science and Technology
  • Huimin Zhao +4
  • Research Article
  • Citations1

Bearing fault diagnosis using transfer learning and optimized deep belief network

  • Feb 09, 2022
  • Measurement Science and Technology
  • Huimin Zhao +4
  • Research Article
  • Citations2

An Enhanced Transfer Learning Remote Sensing Inversion of Coastal Water Quality: A Case Study of Dissolved Oxygen

  • Jan 01, 2025
  • IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
  • Lianwei Li +4
  • Book Chapter
  • Citations7

CaSS: A Channel-Aware Self-supervised Representation Learning Framework for Multivariate Time Series Classification

  • Jan 01, 2022
  • Yijiang Chen +4
  • Research Article
  • Citations1

Self-Supervised Attentive Feature Learning for Alzheimer’s Disease Detection

  • Sep 26, 2025
  • International Journal of Interactive Multimedia and Artificial Intelligence
  • Hela Elmannai +3
  • Research Article
  • Citations12

Sample-distribution-matching-based transfer learning for QoT estimation in optical networks

  • Aug 18, 2023
  • Journal of Optical Communications and Networking
  • Zhiqun Gu +4
  • Research Article
  • Citations4

Transfer learning method for specific emitter identification based on pseudo‐labelling and meta‐learning

  • May 17, 2024
  • IET Radar, Sonar & Navigation
  • Qing Ling +4
  • Research Article
  • Citations1

Contrastive Masked Feature Modeling for Self-Supervised Representation Learning of High-Resolution Remote Sensing Images

  • Feb 17, 2026
  • Remote Sensing
  • Shiyan Pang +4
  • Book Chapter

An Improved Method for Cross-Domain Pedestrian Re-identification

  • Jan 01, 2023
  • Yue Zou +3
  • Conference Article

A Hybrid Transfer Learning Mechanism for Object Classification across View

  • Dec 01, 2012
  • Yi Mo +2
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.