ST3D: Self-training for Unsupervised Domain Adaptation on 3D Object Detection
arXiv:2103.05346
Abstract
We present a new domain adaptive self-training pipeline, named ST3D, for unsupervised domain adaptation on 3D object detection from point clouds. First, we pre-train the 3D detector on the source domain with our proposed random object scaling strategy for mitigating the negative effects of source domain bias. Then, the detector is iteratively improved on the target domain by alternatively conducting two steps, which are the pseudo label updating with the developed quality-aware triplet memory bank and the model training with curriculum data augmentation. These specific designs for 3D object detection enable the detector to be trained with consistent and high-quality pseudo labels and to avoid overfitting to the large number of easy examples in pseudo labeled data. Our ST3D achieves state-of-the-art performance on all evaluated datasets and even surpasses fully supervised results on KITTI 3D object detection benchmark. Code will be available at https://github.com/CVMI-Lab/ST3D.
CVPR2021
References in corpus (19)
- Adam: A Method for Stochastic Optimization
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- Learning Transferable Features with Deep Adaptation Networks
- Deep Learning using Rectified Linear Units (ReLU)
- Conditional Adversarial Domain Adaptation
- FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation
- CyCADA: Cycle-Consistent Adversarial Domain Adaptation
- YOLACT++: Better Real-time Instance Segmentation
- Asymmetric Tri-training for Unsupervised Domain Adaptation
- Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification
- Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID
- Scalability in Perception for Autonomous Driving: Waymo Open Dataset
- Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection
- From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network
- PointDAN: A Multi-Scale 3D Domain Adaption Network for Point Cloud Representation
- Pseudo-Labeling Curriculum for Unsupervised Domain Adaptation
- PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detection
- Complete & Label: A Domain Adaptation Approach to Semantic Segmentation of LiDAR Point Clouds
- Adaptive Semantic Segmentation with a Strategic Curriculum of Proxy Labels