AI Paper Atlas

Research map / Video / Video generation

Beyond Skeletons: Learning Animation Directly from Driving Videos with Same2X Training Strategy

Yuan Zeng, Yujia Shi, Yuhao Yang, Dongxia Liu, Zongqing Lu, Wenming Yang, Qingmin Liao

ICLR 2026 · Video › Video generation · approach: Diffusion-based human image animation · 0 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
add Face Cue (ablation)DirectAnimator · TikTok ablationFIS: 0.418 -> 0.638Improved
add Face Cue (ablation)DirectAnimator · TikTok ablationFVD: 245.48 -> 180.52Improved
enlarge and center the face in Face Cue (ablation)DirectAnimator · TikTok ablationFVD: 185.93 -> 180.52Improved
use gray low-pass-filtered Pose Cue instead of raw RGB (ablation)DirectAnimator · TikTok ablationFVD: 191.92 -> 180.52Improved
add frequency low-pass filtering to gray Pose Cue (ablation)DirectAnimator · TikTok ablationFVD: 184.07 -> 180.52Improved
use cue triplet instead of skeleton maps (ablation)DirectAnimator · TikTok ablationFVD: 216.38 -> 180.52Improved
add soft grid-based Location Cue (ablation)DirectAnimator · TikTok ablationFVD: 203.72 -> 180.52Improved
use soft Location Cue instead of hard mask only (ablation)DirectAnimator · TikTok ablationFVD: 228.54 -> 180.52Improved
add Same2X training strategy (ablation)DirectAnimator · TikTok ablationFVD: 290.43 -> 180.52Improved
increase S2X alignment depth from 10 to 30 (ablation)DirectAnimator · TikTok ablationFVD: 180.52 -> 423.51Hurt
increase S2X alignment depth from 10 to 20 (ablation)DirectAnimator · TikTok ablationFVD: 180.52 -> 230.95Hurt
vary S2X coefficient lambda from 0.5 to 0.1 or 1 (ablation)DirectAnimator · TikTok ablationFVD: 180.52 -> 189.72 / 191.78No clear effect

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