AI Paper Atlas

Research map / Video / Video segmentation and tracking

RoMo: Robust Motion Segmentation Improves Structure from Motion

Lily Goli, Sara Sabour, Mark Matthews, Marcus A. Brubaker, Dmitry Lagun, Alec Jacobson, David J. Fleet, Saurabh Saxena, Andrea Tagliasacchi

ICCV 2025 · Video › Video segmentation and tracking · approach: Learned optical flow estimation · 17 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use RoMo zero-shot masks instead of synthetic-supervised OCLR-adapmotion segmentation · FBMS59IoU: 69.8 -> 75.5Improved
RoMo versus OCLR-adap on single-primary-object VOS benchmarksRoMo · DAVIS16IoU: 80.9 -> 77.3Hurt
add RoMo masks to bundle adjustment on dense tracksTAPIR+TheiaSFM · MPI SintelRPE-R: 1.175 -> 0.217Improved
RoMo masks vs LEAP-VO on Sintel ATERoMo+TheiaSFM · MPI SintelATE: 0.093 vs 0.089No clear effect
add RoMo masks to COLMAPCOLMAP · Casual Motion (8 scenes)ATE: 8.00 -> 2.79Improved
replace SAMv2 features with Stable Diffusion features (ablation)RoMo classifier · SegTrackV2 (13 scenes)IoU: 73.08 -> 71.53No clear effect
add SAMv2 post-processing refinement (ablation)RoMo · SegTrackV2 (13 scenes)IoU: 68.68 -> 73.08Improved
add learned MLP classifier on epipolar labels before SAMv2 (ablation)RoMo · DAVIS2016IoU: 70.9 -> 79.0Improved
increase refinement iterations from 1 to 2 (ablation)RoMo · DAVIS2016IoU: 78.1 -> 79.0Improved
increase refinement iterations from 2 to 3 (ablation)RoMo · DAVIS2016IoU: 79.0 -> 79.0No clear effect
increase MLP hidden size from 8 to 32 (ablation)RoMo classifier · DAVIS2016IoU: 79.0 -> 75.8Hurt
use looser thresholds (theta_u 1.0, theta_l 0.10) (ablation)RoMo · DAVIS2016IoU: 79.0 -> 77.8Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy