Jove
Visualize
Contact Us

Related Concept Videos

State Space Representation01:27

State Space Representation

504
The frequency-domain technique, commonly used in analyzing and designing feedback control systems, is effective for linear, time-invariant systems. However, it falls short when dealing with nonlinear, time-varying, and multiple-input multiple-output systems. The time-domain or state-space approach addresses these limitations by utilizing state variables to construct simultaneous, first-order differential equations, known as state equations, for an nth-order system.
Consider an RLC circuit, a...
504
Classification of Signals01:30

Classification of Signals

1.3K
In signal processing, signals are classified based on various characteristics: continuous-time versus discrete-time, periodic versus aperiodic, analog versus digital, and causal versus noncausal. Each category highlights distinct properties crucial for understanding and manipulating signals.
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
1.3K

You might also read

Related Articles

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

Sort by
Same journal

Human-AI Interaction in Interventional Radiology: A Narrative Review of Current Applications, Challenges, and Future Directions.

Journal of imaging·2026
Same journal

Coronary Artery Anomalies and Anatomical Variants: Cross-Sectional Diagnostic Imaging and Clinical Background.

Journal of imaging·2026
Same journal

YoLeTooth: A Unified Framework for Joint Tooth Segmentation and Periapical Lesion Detection in Panoramic Radiographs.

Journal of imaging·2026
Same journal

Radiomics-Guided Multi-Sequence Learning for Pathological Complete Response Prediction from Breast MRI with Missing Auxiliary Sequences.

Journal of imaging·2026
Same journal

Cutaneous Thermography in Arthropathies: Quantitative Imaging, Machine Learning, and Clinical Translation.

Journal of imaging·2026
Same journal

Two-Stage Dynamic Synergistic Segmentation Method for Myocardial Pathology.

Journal of imaging·2026
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 Video

Updated: Jan 10, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

999

TASA: Text-Anchored State-Space Alignment for Long-Tailed Image Classification.

Long Li1, Tinglei Jia1, Huaizhi Yue1

  • 1School of Information Engineering, Chang'an University, Xi'an 710064, China.

Journal of Imaging
|November 26, 2025
PubMed
Summary
This summary is machine-generated.

This study introduces TASA, a framework improving long-tailed image classification by stabilizing text supervision and enhancing cross-modal fusion. TASA effectively addresses class imbalance using text prototypes and state-space fusion.

Keywords:
cross-modal alignmentlong-tailed image classificationtext prototypesvision–language models

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

727
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

43.5K

Related Experiment Videos

Last Updated: Jan 10, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

999
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

727
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

43.5K

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Long-tailed image classification is difficult for vision-language models due to head class dominance and underrepresented tail classes.
  • Weak text supervision and short prompts exacerbate bias, hindering performance on minority classes.
  • Existing methods struggle with effective cross-modal fusion and stable textual supervision for imbalanced datasets.

Purpose of the Study:

  • To present TASA, an end-to-end framework designed to stabilize textual supervision and enhance cross-modal fusion for long-tailed image classification.
  • To improve the representation of underrepresented tail classes in vision-language models.
  • To develop a method that does not require paired image-text data or per-class prompt tuning.

Main Methods:

  • Implemented a Semantic Distribution Modulation (SDM) module to create stable, diverse class-specific text prototypes using LLM-generated descriptions.
  • Introduced a Dual-Space Cross-Modal Fusion (DCF) module with selective-scan state-space blocks for efficient, bidirectional feature fusion.
  • Utilized a margin-aware alignment loss to align images with class prototypes for classification.

Main Results:

  • TASA demonstrated consistent improvements across many-, medium-, and few-shot groups on CIFAR-10/100-LT, ImageNet-LT, and Places-LT datasets.
  • Ablation studies showed the DCF module provided the largest individual performance gain.
  • The combination of SDM and DCF modules yielded the most robust and balanced classification performance.

Conclusions:

  • Integrating text-driven prototypes with state-space fusion is highly effective for tackling long-tailed image classification challenges.
  • TASA offers a stable and efficient approach to enhance cross-modal fusion and mitigate class bias.
  • The framework successfully aligns visual features with semantic prototypes, improving classification accuracy on imbalanced datasets.