We introduce Timestral, a family of four time series foundation models for probabilistic forecasting, with 5M, 28M, 123M, and 209M parameters. Timestral sets a new state of the art on GIFT-Eval, TIME, and fev-bench, where Timestral-209M and Timestral-123M take the top two places on every metric. We achieve this with th...
Adrien Petralia, A. Gavaudan, Artjom Joosen et al.· 1 citation
Anomaly detection is a fundamental task for time-series analytics, with important implications for the downstream performance of many applications. Despite the large number of anomaly detection methods proposed in the literature, recent benchmark studies have shown that no single detector performs best across highly he...
Christos Panourgias, Roberto Stanzione, Adrien Petralia et al.· 0 citations
LoaDiff is introduced, a diffusion-based generative model for year-long, sub-hourly smart-meter load curves that generates realistic and diverse load profiles, achieves a favorable trade-off between generation quality and limited evidence of memorization, and preserves information useful for downstream energy applicati...
M. Baranova, Adrien Petralia, Étienne Le Naour et al.· arXiv.org· 0 citations
Robostral Navigate, an 8B vision-language model built around this scalability objective, is introduced, which consumes only a stream of monocular RGB images - the most ubiquitous sensor across robotic platforms and predicts waypoints by pointing to the next target location in the current camera view.