AI Paper Atlas

Research map / Video / Video generation

TITAN-Guide: Taming Inference-Time Alignment for Guided Text-to-Video Diffusion Models

Christian Simon, Masato Ishii, Akio Hayakawa, Zhi Zhong, Shusuke Takahashi, Takashi Shibuya, Yuki Mitsufuji

ICCV 2025 · Video › Video generation · approach: Efficient video diffusion models · 2 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
use TITAN-Guide forward-gradient guidance instead of backprop-based guidanceAnimateDiff guided T2V · VGG-Sound audio-video alignmentGPU memory: 67.2GB (DOODL) -> 20.6GBImproved
sample gradient guesses from a subset of frames instead of random Gaussian guesses (ablation)TITAN-Guide · VGG-Sound audio-video alignment (AnimateDiff)FVD: 348.26 -> 331.11Improved
use score-based guesses instead of random guesses (ablation)TITAN-Guide · VGG-Sound audio-video alignment (AnimateDiff)FVD: 348.26 -> 333.42Improved
increase guidance iterations from 20 to 50 or 100 (ablation)TITAN-Guide random guesses · VGG-Sound audio-video alignmentFVD: 348.26 -> 340.21 (50), 350.52 (100)No clear effect
increase resolution from 256x256 to 384x384 (ablation)TITAN-Guide · VGG-Sound audio-video alignment (AnimateDiff)all metrics: FVD 331.11 -> 317.05 (sampled guesses)Improved
traverse to t=0 for clean estimate instead of one-step Tweedie estimateclassifier guidance for T2V · VGG-Sound audio-video alignmentAV-AlignImproved
apply TITAN-Guide on CogVideoX backbone with TFG as baselineCogVideoX · VGG-Sound audio-video alignmentFVD: 343.28 -> 318.99Improved
TITAN-Guide on style guidance ED vs TFGAnimateDiff · TFG-1000 style guidanceED: TFG 8.92 vs TITAN 9.16Hurt
TITAN-Guide on image guidance tasksCAT-DDPM / CelebA-DDPM · Super resolution, CelebA, DeblurringFID/LPIPS/Acc: SR FID 75.9 vs TFG 65.9; CelebA Acc 77.2 vs 75.2Mixed
TITAN-Guide needs extra time to traverse to t=0TITAN-Guide · 16 frames 256x256 on H200guidance time: about 40 seconds -> 2 minutesHurt

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