Annotating Object Instances with a Polygon-RNN

Abstract

We propose an approach for semi-automatic annotation of object instances. While most current methods treat object segmentation as a pixel-labeling problem, we here cast it as a polygon prediction task, mimicking how most current datasets have been annotated. In particular, our approach takes as input an image crop and sequentially produces vertices of the polygon outlining the object. This allows a human annotator to interfere at any time and correct a vertex if needed, producing as accurate segmentation as desired by the annotator. We show that our approach speeds up the annotation process by a factor of 4.7 across all classes in Cityscapes, while achieving 78.4% agreement in IoU with original ground-truth, matching the typical agreement between human annotators. For cars, our speed-up factor is 7.3 for an agreement of 82.2%. We further show generalization capabilities of our approach to unseen datasets.

References in corpus (14)

• Adam: A Method for Stochastic Optimization
• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting
• Deep Residual Learning for Image Recognition
• Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
• The Cityscapes Dataset for Semantic Urban Scene Understanding
• Fully Convolutional Networks for Semantic Segmentation
• Learning to Segment Object Candidates
• Semantic Understanding of Scenes through the ADE20K Dataset
• Pixel-level Encoding and Depth Layering for Instance-level Semantic Labeling
• Monocular Object Instance Segmentation and Depth Ordering with CNNs
• Instance-Level Segmentation for Autonomous Driving with Deep Densely Connected MRFs
• Iterative Instance Segmentation
• DeepCut: Object Segmentation from Bounding Box Annotations using Convolutional Neural Networks

Cited by in corpus (19)

• One-Shot Instance Segmentation
• Learning deep structured active contours end-to-end
• COCO-Stuff: Thing and Stuff Classes in Context
• ModaNet: A Large-Scale Street Fashion Dataset with Polygon Annotations
• Iteratively Trained Interactive Segmentation
• CEREALS - Cost-Effective REgion-based Active Learning for Semantic Segmentation
• Situation Recognition with Graph Neural Networks
• Efficient Interactive Annotation of Segmentation Datasets with Polygon-RNN++
• PolyTransform: Deep Polygon Transformer for Instance Segmentation
• Learning Intelligent Dialogs for Bounding Box Annotation
• Boundary Regularized Building Footprint Extraction From Satellite Images Using Deep Neural Network
• Interactive Medical Image Segmentation with Self-Adaptive Confidence Calibration
• iCap: Interactive Image Captioning with Predictive Text
• CASNet: Common Attribute Support Network for image instance and panoptic segmentation
• One-Click Annotation with Guided Hierarchical Object Detection
• DAGMapper: Learning to Map by Discovering Lane Topology
• Minimizing Labeling Effort for Tree Skeleton Segmentation using an Automated Iterative Training Methodology
• Collaborative Annotation of Semantic Objects in Images with Multi-granularity Supervisions
• ContourRender: Detecting Arbitrary Contour Shape For Instance Segmentation In One Pass