Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Multi-input and Multi-variable systems01:22

Multi-input and Multi-variable systems

441
Cruise control systems in cars are designed as multi-input systems to maintain a driver's desired speed while compensating for external disturbances such as changes in terrain. The block diagram for a cruise control system typically includes two main inputs: the desired speed set by the driver and any external disturbances, such as the incline of the road. By adjusting the engine throttle, the system maintains the vehicle's speed as close to the desired value as possible.
In the absence of...
441
Retrieval01:12

Retrieval

484
Retrieval is the process of getting information out of memory storage and back into conscious awareness. This ability is essential for daily tasks like brushing hair and teeth, driving to work, and performing job duties. Retrieval occurs in three ways: recall, recognition, and relearning.
Recall involves accessing information without cues, such as during an essay test, where individuals must retrieve facts and concepts from memory unaided. Another example is remembering the name of a colleague...
484

You might also read

Related Articles

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

Sort by
Same author

A reconfigurable MTM-EMBG MIMO antenna array with solar panel integration for sustainable 5G networks.

Scientific reports·2026
Same author

Early lumbar drainage plus intrathecal urokinase in severe aneurysmal subarachnoid hemorrhage: the LD-ITUK randomized controlled trial protocol.

Trials·2026
Same author

Analysis of the characteristics of rumen microorganisms and their metabolites and plasma metabolites in crossbred beef cattle at different stages.

Veterinary research communications·2026
Same author

Medical management and revascularization for asymptomatic carotid stenosis: a meta-analysis of randomized controlled trials.

Journal of neurology·2026
Same author

Construction of the ceRNA Regulatory Network Associated with Milk Fat Metabolism.

Animals : an open access journal from MDPI·2026
Same author

Integrating multi-omics to characterize the dynamics of rumen microorganisms and metabolites in Angus cattle at different growth stages.

Research in veterinary science·2026

Related Experiment Video

Updated: Feb 26, 2026

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

5.5K

A privacy-preserving multi-user retrieval system for multimodal artificial intelligence.

Yifang Gao1, Wei Luo2, Chuanchuan Wang3

  • 1School of Electrical and Electronic Engineering, Universiti Sains Malaysia, 14300, Penang, Malaysia.

Scientific Reports
|February 24, 2026
PubMed
Summary
This summary is machine-generated.

This study introduces PMIRS, a secure system for privacy-preserving multimodal retrieval using large language models (LLMs). PMIRS enhances data security through obfuscation, encryption, and federated learning, achieving high accuracy and low latency.

Keywords:
Cross-modal retrievalMultimodal deep learningMultimodal systemsPrivacy-preserving AISearchable encryption

More Related Videos

Author Spotlight: Advancing Large-Scale Neural Dynamics Through HD-MEA Technology
09:44

Author Spotlight: Advancing Large-Scale Neural Dynamics Through HD-MEA Technology

Published on: March 8, 2024

5.9K

Related Experiment Videos

Last Updated: Feb 26, 2026

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

5.5K
Author Spotlight: Advancing Large-Scale Neural Dynamics Through HD-MEA Technology
09:44

Author Spotlight: Advancing Large-Scale Neural Dynamics Through HD-MEA Technology

Published on: March 8, 2024

5.9K

Area of Science:

  • Artificial Intelligence
  • Computer Security
  • Information Retrieval

Background:

  • Large language models (LLMs) are crucial for AI but pose privacy risks in multi-user cloud environments.
  • Secure and private multimodal data retrieval is a significant challenge.

Purpose of the Study:

  • To present PMIRS, a novel system for secure, privacy-focused retrieval of multimodal image and text data.
  • To enable safe transmission of data to cloud-based LLMs while mitigating privacy risks.

Main Methods:

  • PMIRS employs obfuscation techniques, encrypted inference (AES-CBC), and federated learning to fine-tune a CLIP-based model.
  • The Diffie-Hellman algorithm ensures secure key management in multi-user settings.
  • Query embeddings are obfuscated via block-wise projection and encrypted.

Main Results:

  • PMIRS achieved F1-scores up to 0.92 and precision over 0.90 on a customized ImageNet dataset.
  • Retrieval latency was consistently below 180 milliseconds.
  • PMIRS improved average F1-score by 7.67% compared to the CLIP baseline, maintaining precision.

Conclusions:

  • PMIRS offers a practical solution for secure, efficient, and privacy-preserving multimodal retrieval.
  • The system has potential applications in medical imaging, customer service, and enterprise data management under regulations like GDPR and HIPAA.