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

Improving Translational Accuracy02:07

Improving Translational Accuracy

2.5K
2.5K
Vision01:24

Vision

52.2K
Vision is the result of light being detected and transduced into neural signals by the retina of the eye. This information is then further analyzed and interpreted by the brain. First, light enters the front of the eye and is focused by the cornea and lens onto the retina—a thin sheet of neural tissue lining the back of the eye. Because of refraction through the convex lens of the eye, images are projected onto the retina upside-down and reversed.
52.2K
Multi-input and Multi-variable systems01:22

Multi-input and Multi-variable systems

86
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...
86
Force Classification01:22

Force Classification

1.0K
Forces play a crucial role in the study of physics and engineering. They are essential in describing the motion, behavior, and equilibrium of objects in the physical world. Forces can be classified based on their origin, type, and direction of action.
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...
1.0K
Depth Perception and Spatial Vision01:15

Depth Perception and Spatial Vision

453
Depth perception is the ability to perceive objects three-dimensionally. It relies on two types of cues: binocular and monocular. Binocular cues depend on the combination of images from both eyes and how the eyes work together. Since the eyes are in slightly different positions, each eye captures a slightly different image. This disparity between images, known as binocular disparity, helps the brain interpret depth. When the brain compares these images, it determines the distance to an object.
453
Genetic Lingo01:11

Genetic Lingo

98.4K
Overview
98.4K

You might also read

Related Articles

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

Sort by
Same author

Evidence-Based Guidelines for the Diagnosis and Treatment of Pediatric CKD-Mineral and Bone Disorder (Version 2024).

Kidney international reports·2026
Same author

Executive Summary of Evidence-Based Guidelines for the Diagnosis and Treatment of Pediatric CKD-Mineral and Bone Disorder (Version 2024).

Kidney international reports·2026
Same author

The imbalance of Th17/Treg cells exists in asymptomatic hyperuricemia (the early stage of gout).

Clinical rheumatology·2026
Same author

Construction of a nomogram for predicting the risk of primary liver cancer occurrence in patients with liver cysts using CT imaging features combined with alpha-fetoprotein.

Medicine·2025
Same author

A conductive polymer restores connexin43 expression through the suppression of mitogen-activated protein kinases to improve intercellular communication and alleviate atrial fibrillation.

Acta biomaterialia·2025
Same author

Organometallic Iridium-Complex-Functionalized Nanographene Catalysts for Low-Temperature Hydrogenation of Carbonyl Derivatives.

Inorganic chemistry·2024

Related Experiment Video

Updated: May 10, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

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

Published on: December 6, 2024

462

DGPrompt: Dual-guidance prompts generation for vision-language models.

Tai Zheng1, Zhen-Duo Chen1, Zi-Chao Zhang1

  • 1School of Software, Shandong University, 1500 Shunhua Road, Jinan 250101, China.

Neural Networks : the Official Journal of the International Neural Network Society
|April 25, 2025
PubMed
Summary

Dual-Guidance Prompts Generation (DGPrompt) enhances CLIP by improving cross-modal alignment and retaining general knowledge. This method boosts performance in visual recognition tasks, especially with limited data.

Keywords:
Generalization capabilityModalities alignmentPrompt-tuningVision-language model

More Related Videos

Development of a Gaze-Contingent Display Framework Designed for Perceptual and Oculomotor Research with Simulated Central Vision Loss
07:12

Development of a Gaze-Contingent Display Framework Designed for Perceptual and Oculomotor Research with Simulated Central Vision Loss

Published on: April 11, 2025

205
Using the Visual World Paradigm to Study Sentence Comprehension in Mandarin-Speaking Children with Autism
06:15

Using the Visual World Paradigm to Study Sentence Comprehension in Mandarin-Speaking Children with Autism

Published on: October 3, 2018

7.6K

Related Experiment Videos

Last Updated: May 10, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

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

Published on: December 6, 2024

462
Development of a Gaze-Contingent Display Framework Designed for Perceptual and Oculomotor Research with Simulated Central Vision Loss
07:12

Development of a Gaze-Contingent Display Framework Designed for Perceptual and Oculomotor Research with Simulated Central Vision Loss

Published on: April 11, 2025

205
Using the Visual World Paradigm to Study Sentence Comprehension in Mandarin-Speaking Children with Autism
06:15

Using the Visual World Paradigm to Study Sentence Comprehension in Mandarin-Speaking Children with Autism

Published on: October 3, 2018

7.6K

Area of Science:

  • Computer Vision
  • Machine Learning
  • Natural Language Processing

Background:

  • CLIP models achieve strong performance via learnable prompts.
  • Existing methods suffer from limited inter-modal interaction and hierarchical context, hindering alignment.
  • CLIP's sensitivity to prompts leads to overfitting and reduced generalization on unseen classes.

Purpose of the Study:

  • To propose Dual-Guidance Prompts Generation (DGPrompt) for improved visual-textual alignment.
  • To enhance generalization ability by preventing the forgetting of CLIP's general knowledge.
  • To address limitations in existing prompt-tuning methods for CLIP.

Main Methods:

  • DGPrompt utilizes mutual guidance between visual and textual prompts for embedding extraction.
  • A retention module is incorporated to restrain prompt-tuning and preserve general knowledge.
  • The method focuses on aligning visual and textual representation spaces effectively.

Main Results:

  • DGPrompt demonstrates superior performance compared to baseline CLIP and state-of-the-art MaPLe.
  • Achieved an absolute gain of 7.84% and 0.99% on the overall harmonic mean across 11 datasets.
  • Showcased effectiveness in base-to-new class generalization and few-shot learning scenarios.

Conclusions:

  • DGPrompt successfully promotes alignment between visual and textual spaces.
  • The method effectively retains general knowledge, improving generalization on unseen classes.
  • DGPrompt offers a significant advancement in prompt-based learning for vision-language models.