UpCycling: Semi-supervised 3D Object Detection without Sharing Raw-level Unlabeled Scenes
This addresses privacy concerns and computational efficiency for autonomous driving systems by enabling semi-supervised learning without sharing raw data, though it is incremental as it builds on existing SSL and 3D detection methods.
The paper tackles the problem of reducing 3D annotation burden in autonomous driving by proposing UpCycling, a semi-supervised learning framework for 3D object detection that uses unlabeled intermediate features instead of raw point clouds to preserve privacy, achieving performance comparable to or better than state-of-the-art methods that use raw data on datasets like Waymo, KITTI, and Lyft.
Semi-supervised Learning (SSL) has received increasing attention in autonomous driving to reduce the enormous burden of 3D annotation. In this paper, we propose UpCycling, a novel SSL framework for 3D object detection with zero additional raw-level point cloud: learning from unlabeled de-identified intermediate features (i.e., smashed data) to preserve privacy. Since these intermediate features are naturally produced by the inference pipeline, no additional computation is required on autonomous vehicles. However, generating effective consistency loss for unlabeled feature-level scene turns out to be a critical challenge. The latest SSL frameworks for 3D object detection that enforce consistency regularization between different augmentations of an unlabeled raw-point scene become detrimental when applied to intermediate features. To solve the problem, we introduce a novel combination of hybrid pseudo labels and feature-level Ground Truth sampling (F-GT), which safely augments unlabeled multi-type 3D scene features and provides high-quality supervision. We implement UpCycling on two representative 3D object detection models: SECOND-IoU and PV-RCNN. Experiments on widely-used datasets (Waymo, KITTI, and Lyft) verify that UpCycling outperforms other augmentation methods applied at the feature level. In addition, while preserving privacy, UpCycling performs better or comparably to the state-of-the-art methods that utilize raw-level unlabeled data in both domain adaptation and partial-label scenarios.