EUVIMEDOpdag EUVIMED
Open access · CC BY 4.0

IE-MAS: Internal–External Multi-Agent Steering for Controllable Image Captioning

Tiecheng Cai iD, Chao Chen, Shanshan Lin, Sibo Ju iD, Xiangwen Liao

Entropy · 2025-12-07 · 27 (12) · 1237

DOI: 10.3390/e27121237 ↗

At a glance

Journal article in Entropy, published online on 2025-12-07. Written by Tiecheng Cai and 4 co-authors. Institution: College of Computer and Data Science, Fuzhou University, Fuzhou 350108, China. Cited 1 times so far (Crossref); 60 references. The full text is freely available (open access, CC BY 4.0). Funded by: National Natural Science Foundation of China.

AbstractSubjectsBibliografiske oplysningerCiteRelated studies

Abstract (original language)

Controllable Image Captioning (CIC) aims to generate coherent and semantically faithful textual descriptions of images while adhering to user-specified constraints. Existing methods have achieved promising results under individual constraints such as sentimental style or sentence length. However, they typically fail to handle and satisfy multiple constraints simultaneously, as the controls often interact and interfere with one another. To overcome these challenges, we propose Internal–External Multi-Agent Steering (IE-MAS) for CIC. IE-MAS introduces an internal multimodal steering (IMS) strategy to control affective coherence within the caption, and an external multi-agent collaboration system (EMCS) to guide visual grounding and contextual alignment. From an information-theoretic view, IMS reduces uncertainty in the generation process, while EMCS strengthens the dependency between captions and visual inputs, converting the length and sentiment constraints into information gains. Together, they produce a stable balance among semantic consistency, affective expression, and length control through an adaptive steering process that dynamically balances internal linguistic control and external perceptual grounding. Experimental results demonstrate that IE-MAS effectively coordinates multiple constraints, producing captions that satisfy the length constraint and are sentimental expressive and visually faithful.

Subjects

Author keywords

affective analysiscontrollable image captioninginformation theorymulti-agent systems

Authors and institutions

  • College of Computer and Data Science, Fuzhou University, Fuzhou 350108, China.
  • School of Computer Science and Technology, Harbin Institute of Technology (Shenzhen), Shenzhen 518055, China.

ORCID profiles: Tiecheng Cai, Sibo Ju

Funding

  • National Natural Science Foundation of China

Cite

APA

Cai, T., Chen, C., Lin, S., Ju, S., & Liao, X. (2025). IE-MAS: Internal–External Multi-Agent Steering for Controllable Image Captioning. Entropy, 27(12), 1237. https://doi.org/10.3390/e27121237

Vancouver

Cai T, Chen C, Lin S, Ju S, Liao X. IE-MAS: Internal–External Multi-Agent Steering for Controllable Image Captioning. Entropy. 2025;27(12):1237. doi:10.3390/e27121237

BibTeX
@article{Cai2025,
  author = {Cai, Tiecheng and Chen, Chao and Lin, Shanshan and Ju, Sibo and Liao, Xiangwen},
  title = {IE-MAS: Internal–External Multi-Agent Steering for Controllable Image Captioning},
  journal = {Entropy},
  year = {2025},
  volume = {27},
  number = {12},
  pages = {1237},
  doi = {10.3390/e27121237}
}
Start for free →