Identification of shipping signals with few-shot learning: A distribution-aware approach
Bum-Kyu Kim, Sungho Cho, Sunhyo Kim, Hansoo Kim, Ye-Ji Jin, Hyun-Tae Choi, Won-Du Chang · 2026
Read Identification of shipping signals with few-shot learning: A distribution-aware approach by Bum-Kyu Kim, Sungho Cho, Sunhyo Kim, Hansoo Kim, Ye-Ji Jin, Hyun-Tae Choi, Won-Du Chang (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/plos-pone-0352683.txt.
Description
Bum-Kyu Kim and six colleagues present this 2026 PLOS One paper on recognizing ships from underwater sound recordings when only a few examples are available. Navies and researchers often cannot gather large labeled collections for every vessel type, especially where security limits access. The authors build a learning system that can tell ship types apart from just a handful of sample signals.
Their model compares a new recording against stored examples and also weighs how much signals vary within and between ship types. It reached strong accuracy at picking the right class among five options and could handle ship types it had never been trained on. That flexibility matters for ocean monitoring, where new vessels appear and labeled data stay scarce.
Catalog record
| CREATOR | Bum-Kyu Kim, Sungho Cho, Sunhyo Kim, Hansoo Kim, Ye-Ji Jin, Hyun-Tae Choi, Won-Du Chang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35453 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |