Edge-Optimized Semi-Supervised Deep Learning for Power Line Component Inspection †
Nico Surantha, Hanfei Zhang, Daiki Watanabe · 2026
Read Edge-Optimized Semi-Supervised Deep Learning for Power Line Component Inspection † by Nico Surantha, Hanfei Zhang, Daiki Watanabe (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/PMC13363730.txt.
Description
Nico Surantha, Hanfei Zhang, and Daiki Watanabe present a 2026 study on drone-style inspection of power line parts with lean artificial intelligence. Hand labeling of field photos is slow and costly, so their framework learns from a few labeled images plus many unlabeled ones. The framework uses the model's own confident guesses as extra teaching data. A second stage squeezes the model to run in real time on small edge computers.
Their debiased semi-supervised method beat the standard version on detection scores across two datasets. Among hardware trials, a tiny YOLO model in low-precision mode on a Raspberry Pi 5 earned the best balance score on their new deployment scale. They offer that scale as a steadier yardstick for judging edge readiness.
Catalog record
| CREATOR | Nico Surantha, Hanfei Zhang, Daiki Watanabe |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54245 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |