Synthetic Defect Image Generation for Power Line Insulator Inspection Using Multimodal Large Language Models Journal: Engineering Applications of Artificial Intelligence
Xuesong Wang, Caisheng Wang · 2026
Read Synthetic Defect Image Generation for Power Line Insulator Inspection Using Multimodal Large Language Models Journal: Engineering Applications of Artificial Intelligence by Xuesong Wang, Caisheng Wang (2026) 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/arxiv-2603-08069.txt.
Description
Wang and colleagues address a data-scarcity problem in power line insulator inspection, where drone imagery is plentiful but defect examples are rare and inspection sets are often small or closed. Training accurate defect-type classifiers on such thin data is difficult. They use an off-the-shelf multimodal large language model as a training-free image generator, synthesizing defect images from visual references and text prompts. Two-reference conditioning widens variety, light human checks plus prompt tweaks protect label quality, and a filter cleans the synthetic pool.
Catalog record
| CREATOR | Xuesong Wang, Caisheng Wang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52001 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.08069