AI Paper Atlas

Research map / Video / Video generation

GenCompositor: Generative Video Compositing with Diffusion Transformer

Shuzhou Yang, Xiaoyu Li, Xiaodong Cun, Guangzhi Wang, Lingen Li, Ying Shan, Jian Zhang

ICLR 2026 · Video › Video generation · approach: Autoregressive long video generation

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
replace cross-attention injection with DiT fusion block (token concatenation + full self-attention) (ablation)GenCompositor · HYouTubePSNR: 19.8940 -> 42.0010Improved
add background preservation branch (BPBranch) (ablation)GenCompositor · Trajectory-controlled generationBackground consistency: 89.62% -> 93.43%Improved
add BPBranch (ablation)GenCompositor · HYouTubePSNR: 40.0099 -> 42.0010Improved
add luminance (gamma) augmentation of foreground during training (ablation)GenCompositor · HYouTubeLPIPS: 0.0520 -> 0.0385Improved
add mask inflation with Gaussian filter (ablation)GenCompositor · HYouTubePSNR: 41.8553 -> 42.0010Improved
add mask inflation (ablation)GenCompositor · Trajectory-controlled generationBackground consistency: 91.62% -> 93.43%Improved
use ERoPE stream-specific position shift instead of shared RoPE for foreground tokens (ablation)GenCompositor · training losstraining lossImproved
shift ERoPE along height vs width vs time (ablation)ERoPE · training losstraining lossNo clear effect
encode foreground with CLIP instead of VAE under cross-attention (ablation)w/o fusion block variant · qualitativeforeground injectionHurt
GenCompositor vs video harmonization methods Harmonizer and VTTvideo harmonization · HYouTubePSNR: 40.0251 -> 42.0010Improved
GenCompositor vs VACE for trajectory-controlled compositingtrajectory-controlled generation · 40 video setsFVD: 942.52 -> 535.71Improved

Later papers that compared with this method

1 later accepted paper reports GenCompositor as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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