AI Paper Atlas

Research map / 3D vision / 3D generation

ProMo3D: Probing Motion Cues in Frozen 3D Foundation Models

Xiaoang Zhang, Mert Kiray, Yordanka Velikova, Manoj Biswanath, Benjamin Busam

NeurIPS 2026 · 3D vision › 3D generation · approach: Learning 3D reconstruction from images · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Probes whether frozen feed-forward 3D foundation model tokens already encode motion cues that lightweight readouts can expose for dynamic 4D understanding.

Problem

Dynamic 4D understanding with feed-forward 3D foundation models often requires large-scale motion supervision or hand-crafted, architecture-specific heuristics.

Approach

Treats frozen 3D foundation model tokens as the source of motion information and exposes it through lightweight readout functions rather than re-training motion-specific features.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use low-rank quadratic probe instead of linear or diagonal probes (ablation)PMD probes on pi3X, DA3, DUSt3R · DAVIS-all layer-wise probingmotion decodabilityImproved
retrain probe with shuffled token-motion labels (ablation)pi3X rank-8 probe · YouTube-VOS valbalanced accuracy (%): 91.67 -> 50.34Hurt
add SAM2 refinement to pi3X-PMD (ablation)pi3X-PMD · DAVIS17J-M: 73.7 -> 81.9Improved
add SAM2 refinement to pi3-PMD (ablation)pi3-PMD · DAVIS17J-M: 72.7 -> 80.2Improved
use pi3X backbone instead of DUSt3R (ablation)PMD · DAVIS17J-M: 62.2 -> 73.7Improved
use pi3X backbone instead of DA3 (ablation)PMD · DAVIS17J-M: 69.6 -> 73.7Improved
increase probe rank from 8 to 128 (validation) (ablation)pi3X-PMD low-rank probe · YouTube-VOS valnormalized gain x100: 60.77 -> 60.38No clear effect
increase probe rank from 8 to 256 (diagnostic) (ablation)pi3X-PMD low-rank probe · DAVIS-allnormalized gain x100: 66.70 -> 64.66Hurt
use low-rank probe (k=8) rather than full-rank probe with fewer training videos (ablation)pi3X probe · reduced training-video countnormalized gainImproved
reduce number of training videos (ablation)pi3X-PMD · DAVIS-allnormalized gain, J-MHurt
use Otsu + 0.1 threshold instead of default Otsu + 0.2 (ablation)coarse pi3X-PMD · DAVIS-all, SegTrackv2, FBMS-59 meanmean J-M: 72.4 -> 73.0Improved
use extreme fixed thresholds (0.30 or 0.70) (ablation)coarse pi3X-PMD · three-benchmark meanmean J-M: 72.4 -> 66.8 (0.30), 66.4 (0.70)Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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