Active Learning Using Uncertainty Estimation

This repo contains the code related to our paper, Efficacy of Bayesian Neural Networks in Active Learning If you are using this code, please consider citing our work:

Rakesh, V., & Jain, S. (2021). Efficacy of Bayesian Neural Networks in Active Learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 2601-2609).

In our work, we use the following neural networks to estimate uncertainty and consequently perform active learning (AL):

1. Monte Carlo Dropout (MCD)
2. Ensemble-based (EN)
3. Bayesian neural network (BNN)

MCD and EN based codes can be found inside the folder MCD_EN_ActiveLearning and the codes related to BNN can be found inside BNN_ActiveLearning. Readme files can be found inside their respective folders.