ELASTIC: Efficient Once For All Iterative Search for Object Detection on Microcontrollers
Tony Tran, Qin Lin, Bin Hu · 2025
Read ELASTIC: Efficient Once For All Iterative Search for Object Detection on Microcontrollers by Tony Tran, Qin Lin, Bin Hu (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-2503-21999.txt.
Description
Tran, Lin, and Hu present ELASTIC, a search framework that designs object detectors small enough for microcontrollers, in this 2025 paper. Tiny devices leave little room for the bulky parts of a modern detector. Searching all parts at once is too costly, while tuning each part alone misses how they interact.
ELASTIC cycles through the detector's modules, tuning one at a time in rounds. A passthrough trick keeps strong candidates alive between rounds, which steadied the search and sped it up. The gains reached several points of detection accuracy over rival search styles on standard test sets. On real microcontroller chips, the found models used far less energy. The authors share code and videos.
Catalog record
| CREATOR | Tony Tran, Qin Lin, Bin Hu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-21999