Efficient Active Learning for Image Classification and Segmentation Using a Sample Selection and Conditional Generative Adversarial Network

Mahapatra, Dwarikanath; Bozorgtabar, Behzad; Thiran, Jean-Philippe; Reyes, Mauricio (2018). Efficient Active Learning for Image Classification and Segmentation Using a Sample Selection and Conditional Generative Adversarial Network. In: Frangi, Alejandro F.; Schnabel, Julia A.; Davatzikos, Christos; Alberola-López, Carlos; Fichtinger, Gabor (eds.) Medical Image Computing and Computer Assisted Interventions. Lecture Notes in Computer Science: Vol. 11071 (pp. 580-588). Cham: Springer International Publishing 10.1007/978-3-030-00934-2_65

Text MahapatraMiccai2018.pdf - Accepted Version Restricted to registered users only Available under License Publisher holds Copyright. Download (1MB) | Request a copy

Training robust deep learning (DL) systems for medical image classification or segmentation is challenging due to limited images covering different disease types and severity. We propose an active learning (AL) framework to select most informative samples and add to the training data. We use conditional generative adversarial networks (cGANs) to generate realistic chest xray images with different disease characteristics by conditioning its generation on a real image sample. Informative samples to add to the training set are identified using a Bayesian neural network. Experiments show our proposed AL framework is able to achieve state of the art performance by using about $$35\backslash%$$of the full dataset, thus saving significant time and effort over conventional methods.

Interest & Impact

Downloads

2 since deposited on 02 Oct 2019
1 in the past 12 months

Citations

5 Citations in Web of Science ®
6 Citations in Scopus

Search

in Google Scholar™

Services

Actions (login required)

Edit item

Actions (login required)

Edit item