pith. sign in

A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

citation-role summary

dataset 1

citation-polarity summary

fields

cs.CV 1

years

2026 1

verdicts

UNVERDICTED 1

roles

dataset 1

polarities

use dataset 1

representative citing papers

Rethinking Dense Optical Flow without Test-Time Scaling

cs.CV · 2026-05-08 · unverdicted · novelty 6.0

Dense optical flow can be estimated accurately in one forward pass by combining DINO-v2 semantic priors and monocular depth geometric cues via global matching, reaching 2.81 EPE on Sintel Final without any refinement.

citing papers explorer

Showing 1 of 1 citing paper.

  • Rethinking Dense Optical Flow without Test-Time Scaling cs.CV · 2026-05-08 · unverdicted · none · ref 26

    Dense optical flow can be estimated accurately in one forward pass by combining DINO-v2 semantic priors and monocular depth geometric cues via global matching, reaching 2.81 EPE on Sintel Final without any refinement.