A diffusion model-based image generation framework for underwater object detection

Yaoming Zhuang, Longyu Ma, Jiaming Liu, Yonghao Xian, Baoquan Chen, Li Li, Chengdong Wu, Wei Cui, Zhanlin Liu · 2025

Read A diffusion model-based image generation framework for underwater object detection by Yaoming Zhuang, Longyu Ma, Jiaming Liu, Yonghao Xian, Baoquan Chen, Li Li, Chengdong Wu, Wei Cui, Zhanlin Liu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12876851.txt.

Description

Yaoming Zhuang and co-authors use image generation to fix a data drought in underwater vision, in a 2025 paper. Spotting objects under water matters for reef monitoring and rescue work, but real underwater photos are scarce and scenes vary little.

Their framework, ULGF, learns from a small set of underwater images and their labels alone, with no outside photos mixed in. A diffusion model then dreams up fresh, varied underwater scenes that follow the given layouts. The team released the code and the generated data set for others to use. Detection models trained with the synthetic boost did better in real water, and the fake images sat closer to real ones than rival methods managed.

Catalog record

CREATORYaoming Zhuang, Longyu Ma, Jiaming Liu, Yonghao Xian, Baoquan Chen, Li Li, Chengdong Wu, Wei Cui, Zhanlin Liu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE59 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC12876851

Browse the full library · About the Tower of Babel