World2Minecraft: Occupancy-Driven Simulated Scenes Construction
For embodied AI researchers, this provides a customizable simulation platform to address data contamination and flexibility issues, though the approach is incremental.
World2Minecraft converts real-world scenes into structured Minecraft environments using 3D semantic occupancy prediction, enabling downstream tasks like Vision-Language Navigation. They introduce MinecraftOcc, a dataset of 100,165 images from 156 indoor scenes, which improves occupancy prediction and challenges current SOTA methods.
Embodied intelligence requires high-fidelity simulation environments to support perception and decision-making, yet existing platforms often suffer from data contamination and limited flexibility. To mitigate this, we propose World2Minecraft to convert real-world scenes into structured Minecraft environments based on 3D semantic occupancy prediction. In the reconstructed scenes, we can effortlessly perform downstream tasks such as Vision-Language Navigation(VLN). However, we observe that reconstruction quality heavily depends on accurate occupancy prediction, which remains limited by data scarcity and poor generalization in existing models. We introduce a low-cost, automated, and scalable data acquisition pipeline for creating customized occupancy datasets, and demonstrate its effectiveness through MinecraftOcc, a large-scale dataset featuring 100,165 images from 156 richly detailed indoor scenes. Extensive experiments show that our dataset provides a critical complement to existing datasets and poses a significant challenge to current SOTA methods. These findings contribute to improving occupancy prediction and highlight the value of World2Minecraft in providing a customizable and editable platform for personalized embodied AI research. Project page:https://world2minecraft.github.io/.