3D object detection is an important yet demanding task that heavily relies on difficult to obtain 3D annotations. To reduce the required amount of supervision, we propose 3DIoUMatch, a novel semi-supervised method for 3D object detection applicable to both indoor and outdoor scenes. We leverage a teacher-student mutual learning framework to propagate information from the labeled to the unlabeled train set in the form of pseudo-labels. However, due to the high task complexity, we observe that the pseudo-labels suffer from significant noise and are thus not directly usable. To that end, we introduce a confidence-based filtering mechanism, inspired by FixMatch. We set confidence thresholds based upon the predicted objectness and class probability to filter low-quality pseudo-labels. While effective, we observe that these two measures do not sufficiently capture localization quality. We therefore propose to use the estimated 3D IoU as a localization metric and set category-aware self-adjusted thresholds to filter poorly localized proposals. We adopt VoteNet as our backbone detector on indoor datasets while we use PV-RCNN on the autonomous driving dataset, KITTI. Our method consistently improves state-of-the-art methods on both ScanNet and SUN-RGBD benchmarks by significant margins under all label ratios (including fully labeled setting). For example, when training using only 10\% labeled data on ScanNet, 3DIoUMatch achieves 7.7% absolute improvement on mAP@0.25 and 8.5% absolute improvement on mAP@0.5 upon the prior art. On KITTI, we are the first to demonstrate semi-supervised 3D object detection and our method surpasses a fully supervised baseline from 1.8% to 7.6% under different label ratios and categories.
翻译:3D 对象探测是一项重要但又艰巨的任务, 严重依赖难以获得 3D 注释。 为了减少所需的监管量, 我们提议 3DIouUmatch 3D 目标检测的新型半监督性方法, 适用于室内和室外场景。 我们利用教师- 学生相互学习框架, 将标签上的信息传播到假标签形式的无标签列列列中。 然而, 由于任务的复杂性, 我们观察到伪标签受到重大噪音的影响, 因此无法直接使用。 为此, 我们引入了一个基于信任的绝对过滤机制, 由 FixMatch 启发。 我们根据预测对象性和阶级概率概率来筛选低质量的假标签。 我们发现, 这两种措施并不能充分捕捉本地化质量。 我们因此建议使用估计的 3D IoU 作为本地化指标, 并设定有类别自调整的阈值阈值阈值的阈值, 仅用于在内部数据集上使用 PV- RCN 绝对的过滤器 。 我们使用自动驱动器检测器, 3LOIT, 和SAR- bal- dalberal- laveal lader a ex a ex a coal decleg 10 lader a 10 lader a ex a ex ex 10 lader a ex a laveald 10 lab laveald a ex a ex a lader a lader a 10 ex a ex a lader lader a lader a lader a lader laderd laderd laderd a 10 labergmentald labs, 我们 10 a 10 a lader a lader laderd laberd labs a labs a 10 a lab 10 a laberdaldaldaldaldaldaldaldaldaldald 10 a 10 a 10 a 10 a 10 a 10 a 10 a 10 a lad 10 a lad lad lad lad labd 10 a 10 a 10 a 10 a 10 a lad a