← Back to all news
arXiv AI By Alberto D. Cencillo, Leonardo Concepci\'on, Isaac Triguero, Juli\'an Luengo

VACE: Learning Geometrically Structured Representations for Time Series Anomaly Detection

Read the original on arXiv AI →

VACE (Velocity‑Aligned Channel Embeddings) is a self‑supervised anomaly detection method for multivariate time series that learns a geometrically structured representation. It trains a channel‑aware encoder using a velocity‑consistency objective, avoiding negatives and synthetic anomalies, so that normal data form locally smooth, directionally coherent trajectories in embedding space. At inference, VACE combines a Mahalanobis positional score with a directional score that gates the positional score where local dynamics deviate, achieving state‑of‑the‑art performance on the TSB‑AD‑M benchmark.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv Machine Learning
Oct 22

Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection

Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection introduces OracleAD, an unsupervised framework that encodes each variable’s past into a causal embedding to jointly predict the present and reconstruct the input window. The embeddings are projected via self‑attention into a shared latent space, capturing dynamic spatial relationships that are aligned to a Stable Latent Structure (SLS) representing normal-state relationships. Anomalies are detected through a dual scoring mechanism based on prediction error and deviation from the SLS, allowing fine‑grained diagnosis and root‑cause identification at the variable level.

By Dongchan Cho, Jiho Han, Keumyeong Kang, Minsang Kim, Honggyu Ryu, Namsoon Jung
ragdiffusionbenchmarks
More like this →
arXiv Machine Learning
Oct 28

LEAP: a Latent Embedding Anomaly Patch-Transformer for time-series anomaly detection

LEAP is a new time‑series anomaly detection model that avoids reconstruction by using masked prediction and latent‑embedding evidence. It removes patches from a window, predicts them from remaining context, and scores prediction errors as anomaly evidence. The model incorporates three label‑free novelty signals—latent‑embedding novelty, local temporal contrast, and per‑channel novelty—and achieves strong performance on the TSB‑AD benchmark, with statistically significant gains from each component.

By Cristhian Moya-Mota, Ignacio Aguilera-Martos, Diego Garc\'ia-Gil, Juli\'an Luengo
llmsragbenchmarks
More like this →
arXiv AI
Jun 12

ASTER: Latent Pseudo-Anomaly Generation for Unsupervised Time-Series Anomaly Detection

arXiv:2604. 13924v3 Announce Type: replace-cross Abstract: Time-series anomaly detection (TSAD) is critical in domains such as industrial monitoring, healthcare, and cybersecurity, but it remains challenging due to rare and heterogeneous anomalies and the scarcity of labelled data.

By Romain Hermary, Samet Hicsonmez, Dan Pineau, Abd El Rahman Shabayek, Djamila Aouada
llmsragdiffusionbenchmarks
More like this →