AI Paper Atlas

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

UnrealZoo: Enriching Photo-realistic Virtual Worlds for Embodied AI

Fangwei Zhong, Kui Wu, Churan Wang, Hao Chen, Hai Ci, Zhoujun Li, Yizhou Wang

ICCV 2025 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Embodied agent learning · 18 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
parallel rendering, batch commands and IPC sockets (UnrealCV+)UnrealCV · object mask captureFPS: 70 -> 154Improved
UnrealCV+ optimizationsUnrealCV · color image captureFPS: 74 -> 83Improved
UnrealCV+ batch commandsUnrealCV · 10-agent interactionFPS: 8 -> 16Improved
train on 8 environments instead of 1 (ablation)offline RL tracker · 16 unseen environments, Wildssuccess rate: 0.42 -> 0.70Improved
train on 8 environments instead of 1 (ablation)offline RL tracker · TerrainDemosuccess rate: 0.56 -> 0.56No clear effect
increase training diversity from 1 to 8 environments (ablation)offline RL tracker · ModularGothicNightsuccess rate: 0.52 -> 0.76Improved
control frequency from 3 FPS to 30 FPS (ablation)RL agent 1 Env. · FlexibleRoomsuccess rate: 0.34 -> 0.92Improved
add distractors from 0D to 10D (ablation)RL 8 Envs. · social trackingsuccess rate: 0.76 -> 0.56Hurt
add distractors from 0D to 10D (ablation)OpenVLA · social trackingsuccess rate: 0.10 -> 0.00Hurt
transfer from human to robot dog bodyRL 8 Envs. · 4D distractorssuccess rate: 0.68 -> 0.48Hurt
transfer from human to robot dog bodyRL 8 Envs. · 10D distractorssuccess rate: 0.56 -> 0.40Hurt
transfer from human to robot dog bodyPID · 0D distractorssuccess rate: 0.34 -> 0.30Hurt

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