AI Paper Atlas

Research map / Video / Video generation

Captain Cinema: Towards Short Movie Generation

Junfei Xiao, Ceyuan Yang, Lvmin Zhang, Shengqu Cai, Yang Zhao, Yuwei Guo, Gordon Wetzstein, Maneesh Agrawala, Alan Yuille, Lu Jiang

ICLR 2026 · Video › Video generation · approach: Compositional text-to-video diffusion · 40 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 GoldenMem compressed context at 16 pairs (ablation)Captain Cinema keyframe generator · GoldenMem ablation, 16 pairsCompute (PFLOPS est.): 30 -> 21Improved
add GoldenMem at 16 pairs (ablation)Captain Cinema keyframe generator · GoldenMem ablation, 16 pairsVisual consistency: 4.7 -> 4.3Hurt
add GoldenMem at 32 pairs (ablation)Captain Cinema keyframe generator · GoldenMem ablation, 32 pairsCompute (PFLOPS est.): 55 -> 35Improved
add GoldenMem to enable 48 pairs (ablation)Captain Cinema keyframe generator · GoldenMem ablation, 48 pairsfeasibility: OOM -> 52 PFLOPSImproved
add multi-frame interleaved finetuning of video model (ablation)Captain Cinema · Long-video eval (VBench-2.0 style)Temporal dynamic / aesthetic / quality: Aesthetic 56.8 -> 57.2; Quality 60.9 -> 61.7; Dynamic 64.4 -> 65.4Improved
add multi-frame interleaved finetuning of video model (ablation)Captain Cinema · Long-video eval (VBench-2.0 style)Temporal consistency: 91.9 -> 91.0Hurt
use Captain Cinema instead of LCT in long contexts (24 pairs)interleaved long-context generation · Stress test, 24 context pairsIdentity consistency: 0.14 -> 0.42Improved
use Captain Cinema instead of LCT for temporal consistencylong video generation · Long-video eval (VBench-2.0 style)Temporal consistency: 94.8 -> 91.0Hurt
use Captain Cinema instead of LCTlong video generation · User studyvideo quality: 2.4 -> 3.3Improved
progressive long-context finetuning with modest warmup (step 40,000) instead of direct finetuning or late-checkpoint start (ablation)keyframe generator (Flux 1.Dev) · keyframe generationtraining stabilityImproved

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