AI Paper Atlas

Research map / Video / Video generation

The Best of Both Worlds: Integrating Language Models and Diffusion Models for Video Generation

Aoxiong Yin, Xu Tan, Kai Shen, Yichong Leng, Xinyu Zhou, Juncheng Li, Siliang Tang

ICCV 2025 · Video › Video generation · approach: Compositional text-to-video diffusion · 6 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 the semantic video tokenizer (vs ARLON-like VAE-feature LLM) (ablation)LanDiff · VBenchTotal Score: 82.31 -> 85.43Improved
add the semantic video tokenizer (ablation)LanDiff · VBenchSemantic Score: 77.27 -> 82.61Improved
add the semantic video tokenizer (ablation)LanDiff · VBenchQuality Score: 83.58 -> 86.13Improved
add classifier-free guidance for text (ablation)LanDiff · VBenchTotal Score: 81.06 -> 85.43Improved
add classifier-free guidance for text (ablation)LanDiff · VBenchSemantic Score: 73.13 -> 82.61Improved
add classifier-free guidance for text (ablation)LanDiff · VBenchQuality Score: 83.04 -> 86.13Improved
combine LLM semantic tokens with diffusion instead of diffusion alone (same data)LanDiff vs 7B DiT · VBenchTotal Score: 81.85 -> 85.43Improved
combine LLM semantic tokens with diffusion instead of diffusion alone (same data)LanDiff vs 7B DiT · VBenchDynamic Degree: 51.02 -> 92.71Improved
use hybrid LLM+diffusion at 5B vs CogVideoX-5BLanDiff · VBenchTotal Score: 81.61 -> 85.43Improved
use hybrid 5B model vs larger HunyuanVideo 13BLanDiff · VBenchTotal Score: 83.24 -> 85.43Improved
use hybrid model vs ARLON for long videoLanDiff · VBench-LongTotal Score: 65.09 -> 68.34Improved
use LanDiff vs ARLON for long videoLanDiff · VBench-LongDynamic Degree: 50.42 -> 72.86Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 4 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