logo
CanyamAI summaries for academic research
language
MobileMobile
  • Phone FeatureWeChat Mini Program
EnglishEnglish
noticeNotice
BlogBlog
delete
arrow
Return

Implicit visual knowledge enhanced zero-shot image captioning

delete2025-12-17
delete0
PRE
AI
X
Xubo Liu
W
Wenya Guo *
X
Xumeng Liu
R
Ruxue Yan
Y
Ying Zhang
DOI：10.1016/j.eswa.2025.130862delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• We introduce implicit visual knowledge to bridge the gap between CLIP and language models. • A Visual Knowledge Extraction (VKE) module extracts image-related knowledge using GPT-2. • Two integration methods enhance caption accuracy and diversity. • Our framework achieves a 15.9 % improvement in caption diversity on MSCOCO.

Journal

Expert Systems with Applications cover
Expert Systems with Applications
IF:
7.5
Papers:
2.9W
Citations:
10.2W

Organization

No organization information available