Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Oct 14, 2025

Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping

Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping

Published on: December 8, 2023

Cross-attention-map-based regularization for adversarial domain adaptation.

Jingwei Li¹, Huanjie Wang¹, Ke Wu¹

¹Institute of Automation, Chinese Academy of Sciences, Beijing, 100190, China; School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, 100049, China.

Neural Networks : the Official Journal of the International Neural Network Society

|November 4, 2021

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Preemptive Thermochromic Smart Coating for Visual Friction Damage Recognition and Corrosion Protection in Offshore Structures.

Small (Weinheim an der Bergstrasse, Germany)·2026

Same author

Granger-TSllm: Granger causality enhanced LLMs with residual-quantized tokenizer for multivariate time series forecasting.

Neural networks : the official journal of the International Neural Network Society·2025

Same author

An Effective Method To Suppress Side Reactions and Enhance Li<sup>+</sup> Diffusion Rate of LiNi<sub>0.92</sub>Co<sub>0.04</sub>Mn<sub>0.04</sub>O<sub>2</sub> Cathode Material Via Using the Superionic Conductor Li<sub>0.388</sub>Ta<sub>0.238</sub>La<sub>0.475</sub>Cl<sub>3</sub> Covering.

Langmuir : the ACS journal of surfaces and colloids·2025

Same author

One-Step In Situ Synthesis of a Reduced Graphene Oxide-Based Hybrid Hydrogel for Highly Efficient Water Evaporation and Comprehensive Wastewater Treatment.

ACS applied materials & interfaces·2025

Same author

Mental workload recognition from EEG signals via semi-supervised autoencoders.

Computer methods in biomechanics and biomedical engineering·2025

Same author

N<sub>3C</sub>-Defect-Tuned g-C<sub>3</sub>N<sub>4</sub> Photocatalysts: Structural Optimization and Enhanced Tetracycline Degradation Performance.

Nanomaterials (Basel, Switzerland)·2025

Same journal

Anchor-based disentanglement framework for incremental multi-view clustering.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Complex-valued amplitude-phase interference modeling for adversarially robust classification.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

TraNce: Type-aware hypergraph neural network with biological mediators for drug repositioning.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Decentralized ADMM for factorization-based Low-rank matrix estimation.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Memristive neuromorphic circuit design inspired by the neural mechanisms of conditioned fear.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Q-learning based asynchronous Boolean control networks stabilization with data loss.

Neural networks : the official journal of the International Neural Network Society·2026

See all related articles

This study introduces cross-attention-map-based regularization (CAMR) for unsupervised domain adaptation (UDA). CAMR enhances UDA by leveraging few-shot learning (FSL) techniques for finer-grained alignment, improving performance.

Area of Science:

Computer Science
Artificial Intelligence
Machine Learning

Background:

Unsupervised Domain Adaptation (UDA) primarily uses adversarial training to align source and target domains at the distribution or feature level.
Existing UDA methods often overlook finer-grained interactions between classes or samples from different domains.
Few-Shot Learning (FSL) effectively utilizes sample-level interactions, offering a potential avenue for UDA improvement.

Purpose of the Study:

To bridge the gap between Unsupervised Domain Adaptation (UDA) and Few-Shot Learning (FSL) by introducing sample-level alignment strategies into UDA.
To propose a novel method, Cross-Attention-Map-based Regularization (CAMR), for enhancing UDA performance.
To investigate the efficacy of cross-attention mechanisms for regularizing feature maps in UDA.

Keywords:

Attention mechanism Contrastive learning Domain adaptation Few-shot learning

More Related Videos

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Related Experiment Videos

Last Updated: Oct 14, 2025

Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping

Author Spotlight: Insights into Visual Cortex Research Through Wide-View fMRI Mapping

Published on: December 8, 2023

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Main Methods:

High-confidence sample selection (HCSS) to identify relevant samples across domains.
Cross-Attention Map Generation Module (CAMGM) for sample interaction.
Cross-Attention-Map-based Regularization (CAMR) applied to feature extractor outputs.

Main Results:

CAMR demonstrated consistent improvements when integrated with existing UDA objectives.
The proposed method achieved performance gains of 1% to 2% across most tasks on challenging datasets.
CAMR proved effective without requiring additional complex components or extensive tuning.

Conclusions:

Finer-grained sample and class-level interactions, inspired by FSL, can significantly benefit UDA.
CAMR offers a simple yet effective regularization technique for improving UDA.
The approach shows promise for advancing domain adaptation research by incorporating cross-domain sample interactions.