Ultra-Lightweight Network for Ship-Radiated Sound Classification on Embedded Deployment

Sangwon Park, Dongjun Kim, Sung-Hoon Byun, Sangwook Park · 2026

Read Ultra-Lightweight Network for Ship-Radiated Sound Classification on Embedded Deployment by Sangwon Park, Dongjun Kim, Sung-Hoon Byun, Sangwook Park (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-2601-13679.txt.

Description

Sangwon Park, Dongjun Kim, Sung-Hoon Byun, and Sangwook Park build a tiny sound classifier for ships in 2026. Ships give off sounds as they move, and those sounds can reveal the ship type. Small devices at sea must do this job with little power and weak chips. Heavy AI models do not fit that task well.

Their model, ShuffleFAC, uses frequency-aware convolution tuned for low cost. It splits work into light steps and shuffles channels to mix traits cheaply. Tests on the DeepShip dataset showed strong results with far less complexity. Light tools like this suit long, quiet watch at sea.

Catalog record

CREATORSangwon Park, Dongjun Kim, Sung-Hoon Byun, Sangwook Park
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE24 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-13679

Browse the full library · About the Tower of Babel