A low-cost, standalone TinyML device using an ESP32 and two IMUs classifies five gait scenarios with 92% accuracy and produces anomaly scores in under 100 ms.
Advancements in limb prosthetics,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
A Cost-effective, Stand-alone, and Real-time TinyML-Based Gait Diagnosis Unit Aimed at Lower-limb Robotic Prostheses and Exoskeletons
A low-cost, standalone TinyML device using an ESP32 and two IMUs classifies five gait scenarios with 92% accuracy and produces anomaly scores in under 100 ms.