LATIDIA · Robótica

Flujo óptico impulsado por el modelo de cimentación

arXiv:2610.26690v1 Tipo de anuncio: Cross Resumen: Investigamos cómo las representaciones visuales y de profundidad preentrenadas congeladas admiten un flujo óptico denso cuando se combinan con una red de coincidencia supervisada. Basándose en TTSFlow, thi

La noticia

arXiv:2610.26690v1 Announce Type: cross Abstract: We investigate how frozen pretrained visual and depth representations support dense optical flow when coupled with a supervised matching network. Building on TTSFlow, this paper studies Dense Prediction Transformer (DPT) feature decoding and one coarse-to-fine refinement stage. Our method fuses features from frozen DINOv2 and Depth Anything V2 models and estimates optical flow through a trainable transformer-based matching pipeline. After optical-flow training on FlyingChairs and FlyingThings3D, it achieves 2.37 endpoint error (EPE) on the Sintel training split, Final pass, compared with 2.52 for the FlowSeek (T) configuration reported here. Our RoCo-Spring 2026 entry, RoCo-29.VSAI, obtains a robustness-balanced score of 1.1038, Spring EPE of 1.411, and