MMA: A Momentum Mamba Architecture for Human Activity Recognition with Inertial Sensors

TL;DR

This paper introduces Momentum Mamba, an advanced structured state-space model with second-order dynamics that enhances long-term memory, robustness, and efficiency in human activity recognition from inertial sensors.

Contribution

The paper proposes Momentum Mamba, a novel second-order SSM with momentum, improving stability and long-sequence modeling for HAR, and introduces Complex Momentum Mamba for frequency-selective memory.

Findings

Outperforms vanilla Mamba and Transformer baselines in accuracy.

Demonstrates robustness and faster convergence in HAR tasks.

Offers a favorable accuracy-efficiency trade-off.

Abstract

Human activity recognition (HAR) from inertial sensors is essential for ubiquitous computing, mobile health, and ambient intelligence. Conventional deep models such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and transformers have advanced HAR but remain limited by vanishing or exloding gradients, high computational cost, and difficulty in capturing long-range dependencies. Structured state-space models (SSMs) like Mamba address these challenges with linear complexity and effective temporal modeling, yet they are restricted to first-order dynamics without stable longterm memory mechanisms. We introduce Momentum Mamba, a momentum-augmented SSM that incorporates second-order dynamics to improve stability of information flow across time steps, robustness, and long-sequence modeling. Two extensions further expand its capacity: Complex Momentum Mamba for frequency-selective memory scaling. Experiments on multiple HAR benchmarks demonstrate consistent gains over vanilla Mamba and Transformer baselines in accuracy, robustness, and convergence speed. With only moderate increases in training cost, momentum-augmented SSMs offer a favorable accuracy-efficiency balance, establishing them as a scalable paradigm for HAR and a promising principal framework for broader sequence modeling applications.