AI Paper Atlas

Research map / Video / Video generation

Repurposing Video Diffusion Transformers for Cross-View Temporal Object Correspondence

Siyoon Jin, Dahyun Chung, Honggyu An, Sangbeom Lim, Wonjin Nam, Jaewoo Jung, WonJun Moon, Seungryong Kim

NeurIPS 2026 · Video › Video generation · approach: Efficient video diffusion models · 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.

Introduces cross-view temporal object correspondence, which propagates object identity across views and time from a single query mask, and repurposes video diffusion transformers for it.

Problem

Existing cross-view object correspondence methods treat the task as frame-level matching that requires a query mask at every frame, and without temporal context they are ambiguous under appearance changes or occlusions.

Approach

Reframes cross-view correspondence as a temporal problem that preserves object identity over time, and repurposes pretrained video diffusion transformers as the backbone for this task.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add focal cross-view attention alignment loss (ablation)XTrack · Ego-Exo4D Exo2EgoEgo J&F: 0.303 -> 0.332Improved
add hard negative conditioning (ablation)XTrack · Ego-Exo4D Exo2EgoEgo J&F: 0.332 -> 0.431Improved
add local context crop for small objects (ablation)XTrack · Ego-Exo4D Exo2EgoEgo J&F: 0.431 -> 0.509Improved
use pretrained video DiT weights (ablation)XTrack · Ego-Exo4D Exo2EgoEgo J&F: 0.054 -> 0.509Improved
add focal attention alignment loss (ablation)XTrack · Ego-Exo4D Exo2EgoExo J&F: 0.415 -> 0.514Improved
add hard negative conditioning (ablation)XTrack · Ego-Exo4D Exo2EgoExo VA: 0.906 -> 0.882Hurt
joint unified model vs match-then-track with SAM2 (V2-SAM)XTrack vs V2-SAM M->T · Ego-Exo4D Ego2ExoJ&F: 0.545 -> 0.472Hurt
joint view-time prediction instead of decoupled match-then-trackXTrack vs V2-SAM M->T · Ego-Exo4D Ego2ExoVA: 0.628 -> 0.900Improved

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