AI Paper Atlas

Research map / Video / Video generation

TempoControl: Temporal Attention Guidance for Text-to-Video Models

Shira Schiber, Ofir Lindenbaum, Idan Schwartz

CVPR 2026 · Video › Video generation · approach: Compositional text-to-video diffusion · 4 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
apply TempoControl attention steering to Wan2.1 1.3BWan2.1-S · One-object temporal controlTemporal Accuracy: 63.94 -> 83.56Improved
apply TempoControl attention steering to Wan2.1 1.3BWan2.1-S · Two-object temporal controlTemporal Accuracy: 37.50 -> 53.17Improved
apply TempoControl attention steering to Wan2.1 1.3BWan2.1-S · Movement temporal controlTemporal Accuracy: 19.00 -> 54.00Improved
apply TempoControl to CogX-SCogX-S 2B · One-object temporal controlTemporal Accuracy: 58.15 -> 63.75Improved
add explicit timing cues to the text promptWan2.1 · One-object temporal controlTemporal Accuracy: 65.18% -> 63.94%No clear effect
add explicit timing cues to the text promptWan2.1 · One-object temporal controlImaging Quality: 59.99% -> 53.76%Hurt
add Pearson correlation loss alone (ablation)TempoControl (Wan2.1-S) · One-object temporal controlTemporal Accuracy / Imaging Quality: Acc 63.94% -> 81.19%; Img 53.76% -> 50.96%Mixed
add entropy regularization alone (ablation)TempoControl (Wan2.1-S) · One-object temporal controlImaging Quality: 53.76% -> 59.52%Improved
add entropy to correlation loss (ablation)TempoControl (Wan2.1-S) · One-object temporal controlImaging Quality: 50.96% -> 57.60%Improved
add magnitude term to correlation plus entropy (ablation)TempoControl (Wan2.1-S) · One-object temporal controlTemporal Accuracy: 78.38% -> 82.50%Improved
remove explicit timing cues from prompt for Wan2.2Wan2.2 5B · One-object temporal controlImaging Quality / absence accuracy: Img 48.13% -> 57.78%; Abs 87.42 -> 71.25%Mixed
apply TempoControl on VBench multiple objectsWan2.1-S · VBench Multiple ObjectsMultiple Object (GRiT): 74.13% -> 76.37%Improved

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