LATIDIA · Robótica

Un método de pronóstico de etiquetas multimodal para series temporales visuo-motoras aperiódicas

arXiv: 2609.07930v2Tipo de anuncio: reemplazar Resumen: Los modelos de aprendizaje profundo se han aplicado cada vez más a la predicción de series de tiempo (TSF) en los últimos años. Tanto los modelos basados en transformadores como los basados en MLP se han utilizado de manera efectiva

La noticia

arXiv:2609.07930v2 Announce Type: replace Abstract: Deep learning models have been increasingly applied to Time Series Forecasting (TSF) in recent years. Transformer-based and MLP-based models have both been used effectively on many real-world TSF regression benchmarks, and there is ongoing debate as to which family of methods is best. While these benchmarks have drawn much attention, it is also worth noting that many current datasets and methods assume approximate periodicity in the time series. In this work, we focus on a new TSF task without periodicity: anticipating falls during humanoid locomotion, on the basis of egocentric vision and proprioception. When the locomotion trajectories are sufficiently diverse, periodicity is violated. We contribute