Machine learning of hierarchical clustering to segment 2D and 3D images

Abstract

We aim to improve segmentation through the use of machine learning tools during region agglomeration. We propose an active learning approach for performing hierarchical agglomerative segmentation from superpixels. Our method combines multiple features at all scales of the agglomerative process, works for data with an arbitrary number of dimensions, and scales to very large datasets. We advocate the use of variation of information to measure segmentation accuracy, particularly in 3D electron microscopy (EM) images of neural tissue, and using this metric demonstrate an improvement over competing algorithms in EM and natural images.

15 pages, 8 figures

References in corpus (1)

• Natural Scales in Geographical Patterns

Cited by in corpus (25)

• The Mutex Watershed and its Objective: Efficient, Parameter-Free Graph Partitioning
• Image Segmentation Using Hierarchical Merge Tree
• Deep Learning Convolutional Networks for Multiphoton Microscopy Vasculature Segmentation
• A Context-aware Delayed Agglomeration Framework for Electron Microscopy Segmentation
• Self-Supervised Neuron Segmentation with Multi-Agent Reinforcement Learning
• Cylindrical Shape Decomposition for 3D Segmentation of Tubular Objects
• Anisotropic EM Segmentation by 3D Affinity Learning and Agglomeration
• SSHMT: Semi-supervised Hierarchical Merge Tree for Electron Microscopy Image Segmentation
• Combinatorial Energy Learning for Image Segmentation
• Large-Scale Electron Microscopy Image Segmentation in Spark
• Robust neural circuit reconstruction from serial electron microscopy with convolutional recurrent networks
• Learned versus Hand-Designed Feature Representations for 3d Agglomeration
• Learning and Segmenting Dense Voxel Embeddings for 3D Neuron Reconstruction
• Guided Proofreading of Automatic Segmentations for Connectomics
• What Properties are Desirable from an Electron Microscopy Segmentation Algorithm
• Small Sample Learning of Superpixel Classifiers for EM Segmentation- Extended Version
• Scaling Distributed Training of Flood-Filling Networks on HPC Infrastructure for Brain Mapping
• Efficient 2D neuron boundary segmentation with local topological constraints
• An Automated Images-to-Graphs Framework for High Resolution Connectomics
• Multi-stage Multi-recursive-input Fully Convolutional Networks for Neuronal Boundary Detection
• AxonEM Dataset: 3D Axon Instance Segmentation of Brain Cortical Regions
• Convolutional nets for reconstructing neural circuits from brain images acquired by serial section electron microscopy
• SANTIAGO: Spine Association for Neuron Topology Improvement and Graph Optimization
• Icon: An Interactive Approach to Train Deep Neural Networks for Segmentation of Neuronal Structures
• Neural Reconstruction Integrity: A metric for assessing the connectivity of reconstructed neural networks