AI Paper Atlas

Research map / Video / Video generation

Robotic Manipulation by Imitating Generated Videos Without Physical Demonstrations

Shivansh Patel, Shraddhaa Mohan, Hanlin Mai, Unnat Jain, Svetlana Lazebnik, Yunzhu Li

ICLR 2026 · Video › Video generation · approach: World-model video generation

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
filter generated videos with a GPT-4o check before execution (ablation)RIGVid (Kling v1.6) · Sweeping trashsuccess rate: 20% -> 70%Improved
filter generated videos with a VLM (ablation)RIGVid (Kling v1.6) · Placing a spatulasuccess rate: 50% -> 90%Improved
filter generated videos with a VLM (ablation)RIGVid (Kling v1.6) · Pouring watersuccess rate: 80% -> 100%Improved
use better video generator (Sora -> Kling v1.5 -> Kling v1.6) (ablation)RIGVid · four real-world taskssuccess rate: Sora 0% success; Kling v1.6 highestImproved
replace real human demonstration videos with filtered generated videos (ablation)RIGVid · four real-world taskssuccess rateNo clear effect
use 6D object pose trajectories instead of VLM keypoint constraints (ReKep)RIGVid · four real-world taskssuccess rate: 50% -> 85%Improved
use 6D pose tracking instead of point tracks (Gen2Act) (ablation)RIGVid · four real-world taskssuccess rate: 67.5% -> 85.0%Improved
use 6D pose tracking instead of optical flow (AVDC), feature fields (4D-DPM) or goal-image tracks (Track2Act) (ablation)RIGVid · four real-world taskssuccess rate: 32.5% / 35.0% / 7.5% -> 85.0%Improved
score videos with VBench++ video-text or I2V subject consistency instead of VLM filter (ablation)video filtering · four tasksPearson correlation with human judgment: 0.34 / 0.37 vs 0.84Hurt
use monocular generated depth instead of real RGBD depth (ablation)RIGVid · real videossuccess rate: 100% -> 85%Hurt
use MegaPose RGB-only tracking instead of FoundationPose (ablation)RIGVid · pouring trajectoriesrotational RMS jitter: 14.31 -> 37.47 degreesHurt
use mesh-free BundleSDF tracking instead of FoundationPose (ablation)RIGVid · pouringsuccess rate: 90% success; ~30 minutes per videoNo clear effect

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