MicroBi-ConvLSTM: An Ultra-Lightweight Efficient Model for Human Activity Recognition on Resource Constrained Devices
This addresses the need for efficient models for wearable devices with limited memory, though it is incremental as it builds on existing lightweight architectures.
The paper tackles the problem of human activity recognition on resource-constrained wearables by proposing MicroBi-ConvLSTM, an ultra-lightweight model with 11.4K parameters that achieves competitive performance, such as 93.41% macro F1 on UCI-HAR and 88.98% on Daphnet gait freeze detection, while reducing parameters by 2.9x versus prior work.
Human Activity Recognition (HAR) on resource constrained wearables requires models that balance accuracy against strict memory and computational budgets. State of the art lightweight architectures such as TinierHAR (34K parameters) and TinyHAR (55K parameters) achieve strong accuracy, but exceed memory budgets of microcontrollers with limited SRAM once operating system overhead is considered. We present MicroBi-ConvLSTM, an ultra-lightweight convolutional-recurrent architecture achieving 11.4K parameters on average through two stage convolutional feature extraction with 4x temporal pooling and a single bidirectional LSTM layer. This represents 2.9x parameter reduction versus TinierHAR and 11.9x versus DeepConvLSTM while preserving linear O(N) complexity. Evaluation across eight diverse HAR benchmarks shows that MicroBi-ConvLSTM maintains competitive performance within the ultra-lightweight regime: 93.41% macro F1 on UCI-HAR, 94.46% on SKODA assembly gestures, and 88.98% on Daphnet gait freeze detection. Systematic ablation reveals task dependent component contributions where bidirectionality benefits episodic event detection, but provides marginal gains on periodic locomotion. INT8 post training quantization incurs only 0.21% average F1-score degradation, yielding a 23.0 KB average deployment footprint suitable for memory constrained edge devices.