AI Paper Atlas

Research map / Video / Video generation

Bi-directional Autoregressive Diffusion for Large Complex Motion Interpolation

Yongrui Ma, Shijie Zhao, Mingde Yao, Junlin Li, Li Zhang, Xiaohong Liu, Qi Dou, Jinwei Gu, Tianfan Xue

CVPR 2026 · Video › Video generation · approach: Efficient video diffusion models

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
bidirectional autoregressive interpolation instead of full-sequence (ablation)frame interpolation diffusion transformer · FCVG eval setFID: 27.63 -> 23.41Improved
add separate DINOv3 motion estimator on top of Bi-AR (ablation)Frame Bi-AR · FCVG eval setFID: 23.41 -> 19.03Improved
jointly generate flow visualization with frames (ablation)Frame Bi-AR · FCVG eval setFID: 23.41 -> 23.88No clear effect
jointly generate DINOv3 features with frames in one transformer (ablation)Frame Bi-AR · FCVG eval setFID: 23.41 -> 28.97Hurt
bidirectional autoregressive scheme reduces sampling stepsWan · 1024x576 runtimeper-frame runtime: 2.582 -> 0.775 sImproved
use ARVFI vs AMT on DAVIS-7AMT · DAVIS-7LPIPS: 0.254 -> 0.257Hurt
use ARVFI vs WanWan · PixelsFVD: 131.22 -> 101.71Improved

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