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 Experiment Videos

Multisource Port Inspection Sensor Fusion with Causal Representation Learning for Cross-Border Anomaly Monitoring.

Jiaxin Yin1, Zhengjia Lu2,3, Baodi Xiong2

  • 1China Agricultural University, Beijing 100083, China.

Sensors (Basel, Switzerland)
|July 15, 2026
PubMed
Summary

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

Multi-dimensional spatial-temporal projection ultrafast compressed imaging.

Light, science & applications·2026
Same author

Correction: ZNF655 promotes tumor growth and chemoresistance by targeting MAFF-CCND1 axis in ovarian cancer.

Cancer cell international·2026
Same author

Intrinsic electronic structure and nanoscale domains in the kagome superconductor KV3Sb5probed by space resolved photoemission spectroscopy.

Journal of physics. Condensed matter : an Institute of Physics journal·2026
Same author

An integrated multitarget screening strategy combining network pharmacology with cell membrane chromatography coupled to high-performance liquid chromatography for identifying candidate antidepressant compounds from Cyperus rotundus L.

Journal of ethnopharmacology·2026
Same author

Physio-transcriptomic perspectives on midgut structural disruption and metabolic imbalance underlying neodymium oxide toxicity in Bombyx mori.

Toxicology and applied pharmacology·2026
Same author

Adjuvant therapy for renal cell carcinoma: lessons from past failures and new opportunities in the era of immune checkpoint inhibition.

Frontiers in immunology·2026

This study introduces a new framework for detecting cross-border trade anomalies by fusing multisource data. The approach enhances accuracy and interpretability in identifying risks within complex international logistics networks.

Area of Science:

  • Data Science
  • Artificial Intelligence
  • International Trade Security

Background:

  • Cross-border trade generates vast, heterogeneous data from diverse sources, posing challenges for traditional financial review and compliance auditing.
  • Conventional deep learning models struggle with interpretability and can be misled by spurious correlations in complex datasets.

Purpose of the Study:

  • To propose a novel framework for cross-border trade anomaly detection using multisource sensing signal fusion and causal explainability.
  • To overcome limitations in characterizing multisource signal coupling and enhance the interpretability of anomaly detection models.

Main Methods:

  • Uniformly modeling diverse data (texts, financial fields, trajectories, sensor data) as multisource sensing signals.
  • Employing a cross-modal alignment mechanism for collaborative representation of textual semantics, logistics, and device records.
Keywords:
RFID and electronic seal monitoringcold-chain sensor networkscross-border anomaly monitoringexplainable anomaly detectionmultisource sensor data fusion

Related Experiment Videos

  • Implementing an engineering-constraint-guided causal risk module and a counterfactual anomaly response module for explainable risk identification.

    • Main Results:

    • The proposed framework achieved superior performance in cross-border trade anomaly detection, with key metrics like Accuracy (0.927) and AUC (0.958).
    • The model demonstrated robustness across various testing scenarios, including different times, regions, ports, entities, and under challenging conditions like data noise and missing modalities.
    • Ablation studies confirmed the significant contributions of cross-modal attention, causal debiasing, and counterfactual response mechanisms to performance enhancement.

    Conclusions:

    • The developed framework effectively fuses multisource sensing signals for accurate and explainable cross-border trade anomaly detection.
    • The causal and counterfactual components significantly reduce spurious correlations and improve the identification of true anomaly drivers.
    • The method offers a robust solution for enhancing security and compliance in increasingly complex international trade environments.