AI Paper Atlas

Research map / Video / Video generation

GENMAC: Compositional Text-to-Video Generation with Multi-Agent Collaboration

Kaiyi Huang, Yukun Huang, Xuefei Ning, Zinan Lin, Yu Wang, Xihui Liu

AAAI 2026 · Video › Video generation · approach: Compositional text-to-video diffusion · 18 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 GENMAC to VideoCrafter2VideoCrafter2 · T2V-CompBench: Generative numeracyG-Dino score: 0.2869 -> 0.5166Improved
add GENMAC to HunyuanVideoHunyuanVideo · T2V-CompBench: Generative numeracyG-Dino score: 0.1094 -> 0.2109Improved
add GENMAC to HunyuanVideoHunyuanVideo · T2V-CompBench: Motion bindingDOT score: 0.2067 -> 0.2331Improved
add self-routing for correction agents (ablation)GENMAC+VideoCrafter2 · T2V-CompBench averageaverage score: 0.5800 -> 0.6021Improved
add DESIGN stage to generation only (ablation)VideoCrafter2 · T2V-CompBench averageaverage score: 0.4684 -> 0.5154Improved
add REDESIGN stage and iterative refinement (ablation)GENMAC+VideoCrafter2 · T2V-CompBench averageaverage score: 0.5154 -> 0.6021 (full)Improved
use single agent instead of division-of-labor in REDESIGN (ablation)GENMAC+VideoCrafter2 · T2V-CompBench averageaverage score: below 0.6021 fullHurt
remove Suggestion Agent (Verification+Correction only) (ablation)GENMAC · qualitative (Figure 5)correction effectivenessHurt
remove Output Structuring Agent (ablation)GENMAC · qualitative (Figure 5)correction effectiveness: baselines hit max 5 iterations vs 2 for fullHurt
add GENMAC to VideoCrafter2 on object interactionsVideoCrafter2 · T2V-CompBench: Object interactionsGrid-LLaVA score: 0.8125 -> 0.8250No 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