μ \mu Bi-ConvLSTM: An Ultra-Lightweight Efficient Model for Human Activity Recognition on Resource Constrained Devices
Mridankan Mandal · 2026
Read μ \mu Bi-ConvLSTM: An Ultra-Lightweight Efficient Model for Human Activity Recognition on Resource Constrained Devices by Mridankan Mandal (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-2602-06523.txt.
Description
Wearable activity trackers run on microcontrollers with tiny memory, so recognition models must be extremely small. Mridankan Mandal presents MicroBi-ConvLSTM, an ultra-lightweight model built for that regime. Leading lightweight models still exceed microcontroller budgets once the operating system's share is counted.
The design uses two stages of convolutional feature extraction with temporal pooling, followed by a single bidirectional recurrent layer, for an average of just over eleven thousand parameters. Across eight activity benchmarks, it stayed competitive, scoring strongly on daily activities, assembly gestures, and gait freezing detection. On-device tests on two popular boards confirmed it runs within real hardware limits, with quantized integer arithmetic.
Catalog record
| CREATOR | Mridankan Mandal |
|---|---|
| YEAR | 2026 |
| 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/2602-06523