PVP: Polar Representation Boost for 3D Semantic Occupancy Prediction

TL;DR

This paper introduces PVP, a novel polar coordinate-based 3D occupancy prediction method that effectively addresses feature distortion, leading to improved accuracy in 3D semantic occupancy tasks.

Contribution

The paper proposes the Polar Voxel Occupancy Predictor (PVP) with GRP and PD-Conv modules, enhancing polar coordinate representations for 3D occupancy prediction.

Findings

Outperforms existing methods on OpenOccupancy dataset

Achieves higher mIoU and IoU metrics

Effectively handles feature distortion in polar grids

Abstract

Recently, polar coordinate-based representations have shown promise for 3D perceptual tasks. Compared to Cartesian methods, polar grids provide a viable alternative, offering better detail preservation in nearby spaces while covering larger areas. However, they face feature distortion due to non-uniform division. To address these issues, we introduce the Polar Voxel Occupancy Predictor (PVP), a novel 3D multi-modal predictor that operates in polar coordinates. PVP features two key design elements to overcome distortion: a Global Represent Propagation (GRP) module that integrates global spatial data into 3D volumes, and a Plane Decomposed Convolution (PD-Conv) that simplifies 3D distortions into 2D convolutions. These innovations enable PVP to outperform existing methods, achieving significant improvements in mIoU and IoU metrics on the OpenOccupancy dataset.