analiza aiqbench
STREAM-VAE introduces a dual-path variational autoencoder architecture specifically designed to disentangle slow drifts from fast spikes in automotive telemetry, addressing a key limitation of standard single-latent models. By explicitly separating these heterogeneous time scales during reconstruction, the method achieves superior robustness for anomaly detection across diverse vehicle operating modes compared to leading forecasting and attention-based baselines. This advancement offers practical value for both real-time onboard monitoring systems and backend fleet analytics requiring stable performance under dynamic conditions.
STREAM-VAE: Dual-Path Routing for Slow and Fast Dynamics in Vehicle Telemetry Anomaly Detection
arXiv:2511.15339v3 Announce Type: replace Abstract: Automotive telemetry data exhibits slow drifts and fast spikes, often within the same sequence, making reliable anomaly detection challenging. Standard reconstruction-based methods, including sequence variational autoencoders (VAEs), use a single latent process and therefore mix heterogeneous time scales, which can smooth out spikes or inflate variances and weaken anomaly separation. In this paper, we present STREAM-VAE, a variational autoencoder for anomaly detection in automotive telemetry time-series data. Our model uses a dual-path encoder to separate slow drift and fast spike signal dynamics, and a decoder that represents transient deviations separately from the normal operating pattern. STREAM-VAE is designed for deployment, producing stable anomaly scores across operating modes for both in-vehicle monitors and backend fleet analytics. Experiments on an automotive telemetry dataset and the public SMD benchmark show that explicitly separating drift and spike dynamics improves robustness compared to strong forecasting, attention, graph, and VAE baselines.
Abstract:Automotive telemetry data exhibits slow drifts and fast spikes, often within the same sequence, making reliable anomaly detection challenging. Standard reconstruction-based methods, including sequence variational autoencoders (VAEs), use a single latent process and therefore mix heterogeneous time scales, which can smooth out spikes or inflate variances and weaken anomaly separation. In this paper, we present STREAM-VAE, a variational autoencoder for anomaly detection in automotive telemetry time-series data. Our model uses a dual-path encoder to separate slow drift and fast spike signal dynamics, and a decoder that represents transient deviations separately from the normal operating pattern. STREAM-VAE is designed for deployment, producing stable anomaly scores across operating modes for both in-vehicle monitors and backend fleet analytics. Experiments on an automotive telemetry dataset and the public SMD benchmark show that explicitly separating drift and spike dynamics improves robustness compared to strong forecasting, attention, graph, and VAE baselines.
Published in the proceedings of the 2026 IEEE Intelligent Vehicles Symposium (IV 2026), Detroit, MI, USA, June 22-25, 2026. 8 pages, 4 figures, 4 tables
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2511.15339 [cs.LG]
(or arXiv:2511.15339v3 [cs.LG] for this version)
https://doi.org/10.1109/IV66570.2026.11623993 DOI(s) linking to related resources
Text integral sub licența deschisă a publicației · 190 cuvinte
