Disentanglement based Active Learning

Abstract

We propose Disentanglement based Active Learning (DAL), a new active learning technique based on self-supervision which leverages the concept of disentanglement. Instead of requesting labels from human oracle, our method automatically labels the majority of the datapoints, thus drastically reducing the human labeling budget in Generative Adversarial Net (GAN) based active learning approaches. The proposed method uses Information Maximizing Generative Adversarial Nets (InfoGAN) to learn disentangled class category representations. Disagreement between active learner predictions and InfoGAN labels decides if the datapoints need to be human-labeled. We also introduce a label correction mechanism that aims to filter out label noise that occurs due to automatic labeling. Results on three benchmark datasets for the image classification task demonstrate that our method achieves better performance compared to existing GAN-based active learning approaches.

Published in International Joint Conference on Neural Networks (IJCNN), 2021

References in corpus (10)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Conditional Generative Adversarial Nets
• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• Spectral Normalization for Generative Adversarial Networks
• Training Deep Neural Networks on Noisy Labels with Bootstrapping
• Generative Adversarial Active Learning
• Weakly Supervised Disentanglement with Guarantees
• Sampling Strategies for GAN Synthetic Data
• Active Learning from Imperfect Labelers
• Bayesian Generative Active Deep Learning