AI Paper Atlas

Research map / Video / Video segmentation and tracking

Understanding Physical Dynamics with Counterfactual World Modeling

Rahul Venkatesh, Honglin Chen, Kevin Feigelis, Daniel M Bear, Khaled Jedoui, Klemen Kotar, Felix J Binder, Wanhee Lee, Sherry Liu, Kevin Smith, Judith E. Fan, Daniel Yamins

ECCV 2024 · Video › Video segmentation and tracking · approach: Physics-inspired motion prediction · 8 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 temporally-factored masking instead of random tube masking (ablation)CWM · Physion v1.5 OCPaccuracy: 73.2 -> 75.9Improved
add keypoint patch features to classifier input (ablation)CWM · Physion v1.5 OCPaccuracy: 73.6 -> 74.4Improved
add optical flow patches to classifier input (ablation)CWM · Physion v1.5 OCPaccuracy: 74.4 -> 75.5Improved
add segment features to classifier input (ablation)CWM · Physion v1.5 OCPaccuracy: 75.5 -> 75.9Improved
train 1600 epochs instead of 800 (ablation)CWM · Physion v1.5 OCPaccuracy: 75.4 -> 75.9Improved
use 4 context frames instead of 2 during pretraining (ablation)CWM · Physion v1.5 OCPaccuracy: 75.9 -> 68.5Hurt
use patch size 8 instead of 16 (ablation)CWM · Physion v1.5 OCPaccuracy: 74.2 -> 75.9Improved
raise last-frame mask ratio to 0.99 from 0.90 (ablation)CWM · Physion v1.5 OCPaccuracy: 75.4 -> 72.5Hurt
scale CWM from ViT-B to ViT-LCWM · Physion v1.5OCP / OCD accuracy: OCP 75.9 -> 76.1; OCD 89.1 -> 88.7Mixed
apply temporally-factored CWM instead of VideoMAE for counterfactual motionDAVIS counterfactual prediction · DAVISFID: 166.3 -> 25.4Improved
overlay ground-truth segment map on GPT-4V prompt framesGPT-4V · Physion v1.5OCP / OCD accuracy: OCP 52.9 -> 58.3; OCD 54.7 -> 67.5Improved
CWM with 3 frames versus VideoMAE with 16 frames for activity recognitionCWM · Something-Something V2accuracy: 54.2 vs 54.7No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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