Instance-Level Segmentation for Autonomous Driving with Deep Densely Connected MRFs

Abstract

Our aim is to provide a pixel-wise instance-level labeling of a monocular image in the context of autonomous driving. We build on recent work [Zhang et al., ICCV15] that trained a convolutional neural net to predict instance labeling in local image patches, extracted exhaustively in a stride from an image. A simple Markov random field model using several heuristics was then proposed in [Zhang et al., ICCV15] to derive a globally consistent instance labeling of the image. In this paper, we formulate the global labeling problem with a novel densely connected Markov random field and show how to encode various intuitive potentials in a way that is amenable to efficient mean field inference [Krähenbühl et al., NIPS11]. Our potentials encode the compatibility between the global labeling and the patch-level predictions, contrast-sensitive smoothness as well as the fact that separate regions form different instances. Our experiments on the challenging KITTI benchmark [Geiger et al., CVPR12] demonstrate that our method achieves a significant performance boost over the baseline [Zhang et al., ICCV15].

References in corpus (6)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
• Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials
• Fully Connected Deep Structured Networks
• Proposal-free Network for Instance-level Object Segmentation
• Monocular Object Instance Segmentation and Depth Ordering with CNNs

Cited by in corpus (11)

• Visual Question Answering: Datasets, Algorithms, and Future Challenges
• Hybrid Task Cascade for Instance Segmentation
• Learning deep structured active contours end-to-end
• Annotating Object Instances with a Polygon-RNN
• Weakly Supervised Instance Segmentation using Class Peak Response
• Attention-guided Unified Network for Panoptic Segmentation
• InstanceCut: from Edges to Instances with MultiCut
• Object Detection Free Instance Segmentation With Labeling Transformations
• Learning Instance Occlusion for Panoptic Segmentation
• SeGAN: Segmenting and Generating the Invisible
• CASNet: Common Attribute Support Network for image instance and panoptic segmentation