Self-Supervision with Superpixels: Training Few-shot Medical Image Segmentation without Annotation

Published in

European Conference on Computer Vision(2020)

External Links:

Generate Graph

TL;DR

A novel self-supervised FSS framework for medical images in order to eliminate the requirement for annotations during training, and superpixel-based pseudo-labels are generated to provide supervision.

Abstract

Few-shot semantic segmentation (FSS) has great potential for medical imaging applications. Most of the existing FSS techniques require abundant annotated semantic classes for training. However, these methods may not be applicable for medical images due to the lack of annotations. To address this problem we make several contributions: (1) A novel self-supervised FSS framework for medical images in order to eliminate the requirement for annotations during training. Additionally, superpixel-based pseudo-labels are generated to provide supervision; (2) An adaptive local prototype pooling module plugged into prototypical networks, to solve the common challenging foreground-background imbalance problem in medical image segmentation; (3) We demonstrate the general applicability of the proposed approach for medical images using three different tasks: abdominal organ segmentation for CT and MRI, as well as cardiac segmentation for MRI. Our results show that, for medical image segmentation, the proposed method outperforms conventional FSS methods which require manual annotations for training.

Authors

D. Rueckert

14 papers

Chen Chen

2 papers

Huaqi Qiu

3 papers

References81 items

Realistic Adversarial Data Augmentation for MR Image Segmentation

FOAL: Fast Online Adaptive Learning for Cardiac Motion Estimation

Improved Baselines with Momentum Contrastive Learning

Prototype Refinement Network for Few-Shot Segmentation

Weakly Supervised Few-shot Object Segmentation using Co-Attention with Visual and Semantic Inputs

Self-Supervision with Superpixels: Training Few-shot Medical Image Segmentation without Annotation

Published in

European Conference on Computer Vision(2020)

External Links:

Generate Graph

TL;DR

Abstract

Authors

D. Rueckert

14 papers

Chen Chen

2 papers

Huaqi Qiu

3 papers

References81 items

Realistic Adversarial Data Augmentation for MR Image Segmentation

FOAL: Fast Online Adaptive Learning for Cardiac Motion Estimation

Improved Baselines with Momentum Contrastive Learning

Prototype Refinement Network for Few-Shot Segmentation

Weakly Supervised Few-shot Object Segmentation using Co-Attention with Visual and Semantic Inputs

Turkay Kart

1 papers

Weakly Supervised Few-shot Object Segmentation using Co-Attention with Visual and Semantic Embeddings

CHAOS Challenge - Combined (CT-MR) Healthy Abdominal Organ Segmentation

Self-supervised learning for medical image analysis using image context restoration

Differentiable Meta-learning Model for Few-shot Semantic Segmentation

Momentum Contrast for Unsupervised Visual Representation Learning

Domain Generalization via Model-Agnostic Learning of Semantic Features

Self-Supervised Difference Detection for Weakly-Supervised Semantic Segmentation

Pyramid Graph Networks With Connection Attentions for Region-Based One-Shot Semantic Segmentation

SILCO: Show a Few Images, Localize the Common Object

Meta-Learning Initializations for Image Segmentation

Models Genesis: Generic Autodidactic Models for 3D Medical Image Analysis

PANet: Few-Shot Image Semantic Segmentation With Prototype Alignment

Attention-Based Multi-Context Guiding for Few-Shot Semantic Segmentation

Self-Supervised Learning for Cardiac MR Image Segmentation by Anatomical Position Prediction

Data Efficient Unsupervised Domain Adaptation for Cross-Modality Image Segmentation

Boosting Few-Shot Visual Learning With Self-Supervision

Learning Representations by Maximizing Mutual Information Across Views

Deep Compressed Sensing

Prior-Aware Neural Network for Partially-Supervised Multi-Organ Segmentation

Self-Supervised Learning via Conditional Motion Propagation

CANet: Class-Agnostic Segmentation Networks With Iterative Refinement and Attentive Few-Shot Learning

Data Augmentation Using Learned Transformations for One-Shot Medical Image Segmentation

AMP: Adaptive Masked Proxies for Few-Shot Segmentation

'Squeeze & Excite' Guided Few-Shot Segmentation of Volumetric Images

Few-shot 3D Multi-modal Medical Image Segmentation using Generative Adversarial Learning

SG-One: Similarity Guidance Network for One-Shot Semantic Segmentation

nnU-Net: Self-adapting Framework for U-Net-Based Medical Image Segmentation

Invariant Information Clustering for Unsupervised Image Classification and Segmentation

Few-Shot Segmentation Propagation with Guided Networks

TADAM: Task dependent adaptive metric for improved few-shot learning

Unsupervised Feature Learning via Non-parametric Instance Discrimination

Taskonomy: Disentangling Task Transfer Learning

Conditional Networks for Few-Shot Semantic Segmentation

Learning to Compare: Relation Network for Few-Shot Learning

Few-Shot Learning with Graph Neural Networks

Squeeze-and-Excitation Networks

One-Shot Learning for Semantic Segmentation

Multi-task Self-Supervised Visual Learning

Self-supervised Learning for Spinal MRIs

Attention is All you Need

Modeling Relational Data with Graph Convolutional Networks

Prototypical Networks for Few-shot Learning

Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks

Multivariate Mixture Model for Myocardial Segmentation Combining Multi-Source Images

Superpixels: An evaluation of the state-of-the-art

Semi-Supervised Classification with Graph Convolutional Networks

Matching Networks for One Shot Learning

Context Encoders: Feature Learning by Inpainting

Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles

Colorful Image Colorization

Learning Representations for Automatic Colorization

Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning

Unsupervised Visual Representation Learning by Context Prediction

Discriminative Unsupervised Feature Learning with Convolutional Neural Networks

Imaging in population science: cardiovascular magnetic resonance in 100,000 participants of UK Biobank - rationale, challenges and approaches

Entropy rate superpixel segmentation

Multimodal classification of Alzheimer's disease and mild cognitive impairment

PET-guided delineation of radiation therapy treatment volumes: a survey of image segmentation techniques

Statistical shape models for 3D medical image segmentation: A review

Concurrent multimodality image segmentation by active contours for radiotherapy treatment planninga).

One-shot learning of object categories

Adaptive Road Following using Self-Supervised Learning and Reverse Optical Flow

Learning a classification model for segmentation

Prevalence of cerebral white matter lesions in elderly people: a population based magnetic resonance imaging study: the Rotterdam Scan Study

Prevalence of cerebral white matter lesions in elderly people: a population based magnetic resonance imaging study. The Rotterdam Scan Study

Optimal approximations by piecewise smooth functions and associated variational problems

Data augmentation using learned transforms for one-shot medical image segmentation

Few-Shot Semantic Segmentation with Prototype Learning

Miccai multi-atlas labeling beyond the cranial vault

Efficient Graph-Based Image Segmentationを用いた圃場図自動作成手法の検討

One shot learning of simple visual concepts

SLIC Superpixels ?

Automated medical image segmentation techniques

Slic super-pixels. Technical report

Concurrent multimodality image segmentation by active contours for radiotherapy treatment planning.

Current methods in medical image segmentation.

Field of Study

Computer Science

Venue Information

Name

European Conference on Computer Vision

Type

conference

URL

https://link.springer.com/conference/eccv

Alternate Names

ECCV
Eur Conf Comput Vis