Semantic Instance Segmentation via Deep Metric Learning

Abstract

We propose a new method for semantic instance segmentation, by first computing how likely two pixels are to belong to the same object, and then by grouping similar pixels together. Our similarity metric is based on a deep, fully convolutional embedding model. Our grouping method is based on selecting all points that are sufficiently similar to a set of "seed points", chosen from a deep, fully convolutional scoring model. We show competitive results on the Pascal VOC instance segmentation benchmark.

References in corpus (15)

• Deep Residual Learning for Image Recognition
• What makes for effective detection proposals?
• Associative Embedding: End-to-End Learning for Joint Detection and Grouping
• Feature Pyramid Networks for Object Detection
• Simultaneous Detection and Segmentation
• Proposal-free Network for Instance-level Object Segmentation
• Instance-aware Semantic Segmentation via Multi-task Network Cascades
• Instance-sensitive Fully Convolutional Networks
• Fully Convolutional Instance-aware Semantic Segmentation
• Deep Watershed Transform for Instance Segmentation
• Pixel-level Encoding and Depth Layering for Instance-level Semantic Labeling
• InstanceCut: from Edges to Instances with MultiCut
• Reversible Recursive Instance-level Object Segmentation
• End-to-End Instance Segmentation with Recurrent Attention
• FastMask: Segment Multi-scale Object Candidates in One Shot

Cited by in corpus (69)

• YOLACT++: Better Real-time Instance Segmentation
• DeeperLab: Single-Shot Image Parser
• A Single-Shot Arbitrarily-Shaped Text Detector based on Context Attended Multi-Task Learning
• Learning to Measure Change: Fully Convolutional Siamese Metric Networks for Scene Change Detection
• One-Shot Instance Segmentation
• Augmentation for small object detection
• 3D-BEVIS: Bird's-Eye-View Instance Segmentation
• Learning deep structured active contours end-to-end
• A Comprehensive Review of Modern Object Segmentation Approaches
• Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation
• Object Discovery with a Copy-Pasting GAN
• MaskLab: Instance Segmentation by Refining Object Detection with Semantic and Direction Features
• EmbedMask: Embedding Coupling for One-stage Instance Segmentation
• SSAP: Single-Shot Instance Segmentation With Affinity Pyramid
• FEELVOS: Fast End-to-End Embedding Learning for Video Object Segmentation
• 3D Instance Segmentation of MVS Buildings
• Conditional Convolutions for Instance Segmentation
• Online Neural Diarization of Unlimited Numbers of Speakers Using Global and Local Attractors
• Rethinking the Heatmap Regression for Bottom-up Human Pose Estimation
• Anchor Diffusion for Unsupervised Video Object Segmentation
• Learning to Group: A Bottom-Up Framework for 3D Part Discovery in Unseen Categories
• Learning a Spatio-Temporal Embedding for Video Instance Segmentation
• Blazingly Fast Video Object Segmentation with Pixel-Wise Metric Learning
• 3D-FUTURE: 3D Furniture shape with TextURE
• RDSNet: A New Deep Architecture for Reciprocal Object Detection and Instance Segmentation
• Identifying Unknown Instances for Autonomous Driving
• Instance Embedding Transfer to Unsupervised Video Object Segmentation
• GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinement
• SG-Net: Spatial Granularity Network for One-Stage Video Instance Segmentation
• Instance Segmentation with Point Supervision
• Spatial Semantic Embedding Network: Fast 3D Instance Segmentation with Deep Metric Learning
• BANet: Bidirectional Aggregation Network with Occlusion Handling for Panoptic Segmentation
• An Error Detection and Correction Framework for Connectomics
• Learning Deep Representations for Semantic Image Parsing: a Comprehensive Overview
• Boundary-preserving Mask R-CNN
• Point Cloud Instance Segmentation using Probabilistic Embeddings
• Memory Aggregation Networks for Efficient Interactive Video Object Segmentation
• Semi-convolutional Operators for Instance Segmentation
• Object Discovery in Videos as Foreground Motion Clustering
• OccuSeg: Occupancy-aware 3D Instance Segmentation
• Learning and Segmenting Dense Voxel Embeddings for 3D Neuron Reconstruction
• Instance and Panoptic Segmentation Using Conditional Convolutions
• Deep Variational Instance Segmentation
• Geometry-Aware Instance Segmentation with Disparity Maps
• Learning About Objects by Learning to Interact with Them
• A^2-FPN: Attention Aggregation based Feature Pyramid Network for Instance Segmentation
• Efficient Full Image Interactive Segmentation by Leveraging Within-image Appearance Similarity
• CASNet: Common Attribute Support Network for image instance and panoptic segmentation
• Improving Image co-segmentation via Deep Metric Learning
• Attend to Who You Are: Supervising Self-Attention for Keypoint Detection and Instance-Aware Association
• Flow Based Self-supervised Pixel Embedding for Image Segmentation
• Towards Neural Diarization for Unlimited Numbers of Speakers Using Global and Local Attractors
• Ro-SOS: Metric Expression Network (MEnet) for Robust Salient Object Segmentation
• Semantic Segmentation and Object Detection Towards Instance Segmentation: Breast Tumor Identification
• Transformer Assisted Convolutional Network for Cell Instance Segmentation
• Single-Image Piece-wise Planar 3D Reconstruction via Associative Embedding
• Human-centric Relation Segmentation: Dataset and Solution
• Bounding Box Embedding for Single Shot Person Instance Segmentation
• VideoClick: Video Object Segmentation with a Single Click
• Instance Segmentation of Biomedical Images with an Object-aware Embedding Learned with Local Constraints
• Unsupervised Discovery of the Long-Tail in Instance Segmentation Using Hierarchical Self-Supervision
• Indoor Panorama Planar 3D Reconstruction via Divide and Conquer
• Layered Embeddings for Amodal Instance Segmentation
• MaskPlus: Improving Mask Generation for Instance Segmentation
• Affinity Derivation and Graph Merge for Instance Segmentation
• Convolutional nets for reconstructing neural circuits from brain images acquired by serial section electron microscopy
• Fast Pixel-Matching for Video Object Segmentation
• A Dataset for Lane Instance Segmentation in Urban Environments
• Semantic Attention and Scale Complementary Network for Instance Segmentation in Remote Sensing Images