Unsupervised Object Detection with LiDAR Clues

Abstract

Despite the importance of unsupervised object detection, to the best of our knowledge, there is no previous work addressing this problem. One main issue, widely known to the community, is that object boundaries derived only from 2D image appearance are ambiguous and unreliable. To address this, we exploit LiDAR clues to aid unsupervised object detection. By exploiting the 3D scene structure, the issue of localization can be considerably mitigated. We further identify another major issue, seldom noticed by the community, that the long-tailed and open-ended (sub-)category distribution should be accommodated. In this paper, we present the first practical method for unsupervised object detection with the aid of LiDAR clues. In our approach, candidate object segments based on 3D point clouds are firstly generated. Then, an iterative segment labeling process is conducted to assign segment labels and to train a segment labeling network, which is based on features from both 2D images and 3D point clouds. The labeling process is carefully designed so as to mitigate the issue of long-tailed and open-ended distribution. The final segment labels are set as pseudo annotations for object detection network training. Extensive experiments on the large-scale Waymo Open dataset suggest that the derived unsupervised object detection method achieves reasonable accuracy compared with that of strong supervision within the LiDAR visible range. Code shall be released.

References in corpus (21)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• A Simple Framework for Contrastive Learning of Visual Representations
• Improved Baselines with Momentum Contrastive Learning
• Multiscale Combinatorial Grouping for Image Segmentation and Object Proposal Generation
• Contrastive Multiview Coding
• Weakly Supervised Object Localization with Multi-fold Multiple Instance Learning
• Deep Continuous Fusion for Multi-Sensor 3D Object Detection
• The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scale
• Vehicle Detection from 3D Lidar Using Fully Convolutional Network
• HDNET: Exploiting HD Maps for 3D Object Detection
• Scalability in Perception for Autonomous Driving: Waymo Open Dataset
• LSDA: Large Scale Detection Through Adaptation
• Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection
• IPOD: Intensive Point-based Object Detector for Point Cloud
• GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose
• Self-labelling via simultaneous clustering and representation learning
• Soft Sampling for Robust Object Detection
• Deep Learning for 3D Point Clouds: A Survey
• Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networks
• Proposal Learning for Semi-Supervised Object Detection