MobileOcc: A Human-Aware Semantic Occupancy Dataset for Mobile Robots

TL;DR

MobileOcc introduces a new dataset and benchmark for dense 3D semantic occupancy perception in pedestrian-rich environments, emphasizing human occupancy modeling and pedestrian velocity prediction for mobile robots.

Contribution

We present MobileOcc, a novel dataset with an annotation pipeline and mesh optimization framework for human occupancy, enabling new benchmarks in occupancy and pedestrian velocity prediction.

Findings

Our dataset supports robust occupancy prediction benchmarks.

The mesh optimization improves human geometry reconstruction accuracy.

Baseline methods demonstrate competitive performance across tasks.

Abstract

Dense 3D semantic occupancy perception is critical for mobile robots operating in pedestrian-rich environments, yet it remains underexplored compared to its application in autonomous driving. To address this gap, we present MobileOcc, a semantic occupancy dataset for mobile robots operating in crowded human environments. Our dataset is built using an annotation pipeline that incorporates static object occupancy annotations and a novel mesh optimization framework explicitly designed for human occupancy modeling. It reconstructs deformable human geometry from 2D images and subsequently refines and optimizes it using associated LiDAR point data. Using MobileOcc, we establish benchmarks for two tasks, i) Occupancy prediction and ii) Pedestrian velocity prediction, using different methods including monocular, stereo, and panoptic occupancy, with metrics and baseline implementations for reproducible comparison. Beyond occupancy prediction, we further assess our annotation method on 3D human pose estimation datasets. Results demonstrate that our method exhibits robust performance across different datasets.