AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

ThreeDWorld: A Platform for Interactive Multi-Modal Physical Simulation

NeurIPS 2021 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Embodied agent learning · 127 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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 TDW-generated images instead of SceneNet images for pretrainingResNet-50 visual representation · Fine-grained image classification (mean)accuracy: 0.40 -> 0.76Helped
Use TDW-generated images instead of AI2-THOR images for pretrainingResNet-50 visual representation · Fine-grained image classification (mean)accuracy: 0.63 -> 0.76Helped
Train on TDW audio instead of Sound-20K audioVGG-16 material classifier · Real-world material recognitionaccuracy: 0.34 -> 0.66Helped
Add audio features to vision-only features2-layer MLP classifier · Material classificationaccuracy: 0.72 -> 0.96Helped
Add audio features to vision-only features2-layer MLP classifier · Mass classificationaccuracy: 0.42 -> 0.83Helped
Replace HRN with DRHRNPhysical dynamics prediction · Lift, Slide, Collide, Stack, and Clothglobal and local position MSEHelped
Remove collision detection from HRN (ablation)HRN physical dynamics predictor · Physical predictionprediction accuracyHurt

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