Cost-Effective Active Learning for Melanoma Segmentation

Abstract

We propose a novel Active Learning framework capable to train effectively a convolutional neural network for semantic segmentation of medical imaging, with a limited amount of training labeled data. Our contribution is a practical Cost-Effective Active Learning approach using dropout at test time as Monte Carlo sampling to model the pixel-wise uncertainty and to analyze the image information to improve the training performance. The source code of this project is available at https://marc-gorriz.github.io/CEAL-Medical-Image-Segmentation/ .

NIPS ML4H 2017 workshop

References in corpus (2)

• Learning Deconvolution Network for Semantic Segmentation
• Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inference

Cited by in corpus (13)

• Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentation
• DSAL: Deeply Supervised Active Learning from Strong and Weak Labelers for Biomedical Image Segmentation
• O-MedAL: Online Active Deep Learning for Medical Image Analysis
• A Survey on Deep Learning of Small Sample in Biomedical Image Analysis
• Uncertainty Quantification in Deep Learning for Safer Neuroimage Enhancement
• Deep Active Learning for Computer Vision: Past and Future
• Self-supervised Assisted Active Learning for Skin Lesion Segmentation
• Minimax Active Learning
• Reducing Annotating Load: Active Learning with Synthetic Images in Surgical Instrument Segmentation
• Deep Active Learning for Efficient Training of a LiDAR 3D Object Detector
• DEAL: Difficulty-aware Active Learning for Semantic Segmentation
• State-Relabeling Adversarial Active Learning
• Mask-guided sample selection for Semi-Supervised Instance Segmentation