Sign Language Recognition using Parallel Bidirectional Reservoir Computing
Nitin Kumar Singh, Arie Rachmad Syulistyo, Yuichiro Tanaka, Hakaru Tamukoh · 2025
Read Sign Language Recognition using Parallel Bidirectional Reservoir Computing by Nitin Kumar Singh, Arie Rachmad Syulistyo, Yuichiro Tanaka, Hakaru Tamukoh (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/arxiv-2512-19451.txt.
Description
Nitin Kumar Singh, Arie Rachmad Syulistyo, Yuichiro Tanaka, and Hakaru Tamukoh recognize sign language with a computing style built for small devices. Deep learning sign readers work well but eat far too much compute for edge hardware. Their system pairs MediaPipe hand tracking with parallel bidirectional reservoir computing.
MediaPipe extracts hand joint coordinates in real time, and two bidirectional echo-state reservoirs, run in parallel, turn the motion stream into rich features for a classifier. On the word-level American Sign Language video set, scores reached about sixty-one percent top-one accuracy. Training took under twenty seconds, against nearly an hour for deep rivals.
Catalog record
| CREATOR | Nitin Kumar Singh, Arie Rachmad Syulistyo, Yuichiro Tanaka, Hakaru Tamukoh |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-19451