ADLGen: Synthesizing Symbolic, Event-Triggered Sensor Sequences for Human Activity Modeling

TL;DR

ADLGen is a novel generative framework that synthesizes realistic, semantically rich sensor sequences for human activity modeling, addressing privacy and data scarcity issues in ambient assistive environments.

Contribution

It introduces a Transformer-based generative model with symbolic encoding and an LLM-enhanced refinement loop for high-fidelity activity data synthesis.

Findings

Outperforms baseline generators in statistical fidelity

Achieves higher semantic richness in generated sequences

Improves downstream activity recognition accuracy

Abstract

Real world collection of Activities of Daily Living data is challenging due to privacy concerns, costly deployment and labeling, and the inherent sparsity and imbalance of human behavior. We present ADLGen, a generative framework specifically designed to synthesize realistic, event triggered, and symbolic sensor sequences for ambient assistive environments. ADLGen integrates a decoder only Transformer with sign based symbolic temporal encoding, and a context and layout aware sampling mechanism to guide generation toward semantically rich and physically plausible sensor event sequences. To enhance semantic fidelity and correct structural inconsistencies, we further incorporate a large language model into an automatic generate evaluate refine loop, which verifies logical, behavioral, and temporal coherence and generates correction rules without manual intervention or environment specific tuning. Through comprehensive experiments with novel evaluation metrics, ADLGen is shown to outperform baseline generators in statistical fidelity, semantic richness, and downstream activity recognition, offering a scalable and privacy-preserving solution for ADL data synthesis.