AI Paper Atlas

Research map / Video / Video generation

Point Prompting: Counterfactual Tracking with Video Diffusion Models

Ayush Shrivastava, Sanyam Mehta, Daniel Geng, Andrew Owens

ICLR 2026 · Video › Video generation · approach: Efficient video diffusion models

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 the unedited first frame as a negative prompt (counterfactual enhancement guidance) (ablation)Point Prompting · TAP-Vid DAVIStrackingImproved
add inpainting-based coarse-to-fine refinement (ablation)Point Prompting · TAP-Vid DAVISpositional accuracyImproved
add color rebalancing to suppress red hues (ablation)Point Prompting · TAP-Vid DAVIStrackingImproved
add local search window to tracker (ablation)Point Prompting tracker · TAP-Vid DAVISAJ: 35.80 -> 38.92Improved
average positions of nearby red pixels (ablation)Point Prompting tracker · TAP-Vid DAVISAJ: 39.08 -> 42.70Improved
use LAB instead of HSV color space (ablation)Point Prompting tracker · TAP-Vid DAVISAJ: 42.70 -> 42.30Hurt
scale video model from Wan2.1-1.3B to 14B (ablation)Point Prompting · TAP-Vid DAVISAJ: 44.58 -> 48.60Improved
upsample low-resolution inputs before generation (ablation)Point Prompting · TAP-Vid KubricAJ: 31.51 -> 33.55Improved
run on original high-resolution DAVIS frames (ablation)Point Prompting · TAP-Vid DAVISAJ: 42.21 -> 48.6Improved
set marker color to red versus other colors (ablation)Point Prompting · TAP-Vid DAVIS subsettrackingNo clear effect
use guidance weight lambda = 8 (ablation)Point Prompting · TAP-Vid DAVIS subsettrackingImproved
distill generator tracks into CoTrackerCoTracker trained from scratch · TAP-Vid DAVISAJ: teacher 38.78 vs distilled 37.17Improved

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