跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 13 天前AI 評分37

LO-World:以統一 BEV 潛在空間聯合預測 4D 佔用與 LiDAR

LO-World: Unified BEV Latent Modeling for Joint 4D Occupancy and LiDAR Forecasting

AI 導讀

LO-World 是一個在共享 BEV 潛在空間中聯合預測 LiDAR 點雲與 3D 語義佔用的統一 4D 世界模型。實驗中,其佔用預測 IoU 提升 10.9%,LiDAR Chamfer Distance 降低 30.1%。模型在單張 RTX 4090 上的即時吞吐量超過 25 FPS。

正文

Published on Nov 9

Authors:

,

,

Abstract

A unified 4D world model jointly forecasts LiDAR point clouds and 3D semantic occupancy in a shared BEV latent space, improving accuracy and real-time throughput.

LiDAR point clouds and 3D semantic occupancy represent two distinct modalities capturing fine-grained metric geometry and dense volumetric semantics, respectively, holding natural potential for mutual complementarity. However, existing world models typically forecast these modalities independently or cascade them heuristically, missing the opportunity for reciprocal cross-modal regularization. To address this, we propose LO-World, a unified 4D world model that forecasts both modalities within a shared BEV latent space. This single-stream formulation enables dense semantics and metric geometry to mutually reinforce each other, while eliminating multi-pipeline redundancy. Specifically, LO-World couples an efficient hybrid range-voxel autoencoder with a flow-matching generator that caches historical context to reduce iterative sampling latency. Extensive experiments demonstrate that LO-World outperforms prior state-of-the-art baselines, improving occupancy forecasting IoU by 10.9% and reducing LiDAR Chamfer Distance by 30.1%. Meanwhile, it sustains a real-time throughput exceeding 25~FPS on a single RTX 4090. Our code is available at https://github.com/lzzzzzm/LO-World.

View arXiv page View PDF Project page GitHub Add to collection

Get this paper in your agent:

hf papers read 2611.11912

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 1

JimmyLiaoo/LO-World

Updated 2 days ago

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2611.11912 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2611.11912 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co