Identifying Unknown Instances for Autonomous Driving

Abstract

In the past few years, we have seen great progress in perception algorithms, particular through the use of deep learning. However, most existing approaches focus on a few categories of interest, which represent only a small fraction of the potential categories that robots need to handle in the real-world. Thus, identifying objects from unknown classes remains a challenging yet crucial task. In this paper, we develop a novel open-set instance segmentation algorithm for point clouds which can segment objects from both known and unknown classes in a holistic way. Our method uses a deep convolutional neural network to project points into a category-agnostic embedding space in which they can be clustered into instances irrespective of their semantics. Experiments on two large-scale self-driving datasets validate the effectiveness of our proposed method.

3rd Conference on Robot Learning (CoRL 2019)

References in corpus (10)

• Prototypical Networks for Few-shot Learning
• Zero-Shot Learning Through Cross-Modal Transfer
• Learning to Segment Object Candidates
• Semantic Instance Segmentation with a Discriminative Loss Function
• HDNET: Exploiting HD Maps for 3D Object Detection
• Semantic Instance Segmentation via Deep Metric Learning
• MASC: Multi-scale Affinity with Sparse Convolution for 3D Instance Segmentation
• 3D-BEVIS: Bird's-Eye-View Instance Segmentation
• TensorMask: A Foundation for Dense Object Segmentation
• Bayesian Semantic Instance Segmentation in Open Set World

Cited by in corpus (10)

• Deep Learning for Image and Point Cloud Fusion in Autonomous Driving: A Review
• Center-based 3D Object Detection and Tracking
• An Application-Driven Conceptualization of Corner Cases for Perception in Highly Automated Driving
• Domain Adaptation Through Task Distillation
• Lidar Panoptic Segmentation in an Open World
• Patchwork: Concentric Zone-based Region-wise Ground Segmentation with Ground Likelihood Estimation Using a 3D LiDAR Sensor
• Synthetic Unknown Class Learning for Learning Unknowns
• RaidaR: A Rich Annotated Image Dataset of Rainy Street Scenes
• Opening up Open-World Tracking
• Video Class Agnostic Segmentation with Contrastive Learning for Autonomous Driving