Neuromorphic visual attention for Sign-language recognition on SpiNNaker
Šárka Lísková, Olha Vedmedenko, Mazdak Fatahi, Matej Hoffmann, P. Michael Furlong, Giulia D’Angelo · 2026
Read Neuromorphic visual attention for Sign-language recognition on SpiNNaker by Šárka Lísková, Olha Vedmedenko, Mazdak Fatahi, Matej Hoffmann, P. Michael Furlong, Giulia D’Angelo (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-2605-06005.txt.
Description
Sarka Liskova and five co-authors build a sign language recognition system that runs on neuromorphic hardware. Recent recognizers are accurate, but their delay and power needs block real time use on small devices. Neuromorphic chips compute with sparse spikes, which suits fast, low power perception at the edge.
Their system pairs a spiking visual attention step, which picks out the region of interest online, with a compact spiking network on the SpiNNaker platform. It targets American Sign Language fingerspelling and was tested on two event based data sets. Accuracy reached 92.27 percent in simulation and 83.1 percent on the hardware itself. The design used just 0.565 milliwatts with a three millisecond delay, the leanest figures among the compared approaches.
Catalog record
| CREATOR | Šárka Lísková, Olha Vedmedenko, Mazdak Fatahi, Matej Hoffmann, P. Michael Furlong, Giulia D’Angelo |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 33 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.06005