SuperQuadricOcc: Real-Time Self-Supervised Semantic Occupancy Estimation with Superquadric Volume Rendering

TL;DR

SuperQuadricOcc introduces a real-time, self-supervised semantic occupancy estimation method using superquadrics, with an efficient volume renderer enabling scalable scene understanding for autonomous driving.

Contribution

It is the first to leverage superquadrics in a self-supervised occupancy model with a novel real-time rendering technique that reduces computational costs.

Findings

Achieves state-of-the-art performance on Occ3D-nuScenes dataset.

Operates at real-time inference speeds with lower memory usage.

Uses fewer primitives than previous Gaussian-based methods.

Abstract

Self-supervision for semantic occupancy estimation is appealing as it removes the labour-intensive manual annotation, thus allowing one to scale to larger autonomous driving datasets. Superquadrics offer an expressive shape family very suitable for this task, yet their deployment in a self-supervised setting has been hindered by the lack of efficient rendering methods to bridge the 3D scene representation and 2D training pseudo-labels. To address this, we introduce SuperQuadricOcc, the first self-supervised occupancy model to leverage superquadrics for scene representation. To overcome the rendering limitation, we propose a real-time volume renderer that preserves the fidelity of the superquadric shape during rendering. It relies on spatial superquadric-voxel indexing, restricting each ray sample to query only nearby superquadrics, thereby greatly reducing memory usage and computational cost. Using drastically fewer primitives than previous Gaussian-based methods, SuperQuadricOcc achieves state-of-the-art performance on the Occ3D-nuScenes dataset, while running at real-time inference speeds with substantially reduced memory footprint.