Episode-Based Active Learning with Bayesian Neural Networks

03/21/2017
by   Feras Dayoub, et al.
0

We investigate different strategies for active learning with Bayesian deep neural networks. We focus our analysis on scenarios where new, unlabeled data is obtained episodically, such as commonly encountered in mobile robotics applications. An evaluation of different strategies for acquisition, updating, and final training on the CIFAR-10 dataset shows that incremental network updates with final training on the accumulated acquisition set are essential for best performance, while limiting the amount of required human labeling labor.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
11/19/2019

Rethinking deep active learning: Using unlabeled data at model training

Active learning typically focuses on training a model on few labeled exa...
research
10/12/2022

Fast Bayesian Updates for Deep Learning with a Use Case in Active Learning

Retraining deep neural networks when new data arrives is typically compu...
research
06/27/2019

Deep Active Learning with Adaptive Acquisition

Model selection is treated as a standard performance boosting step in ma...
research
04/15/2021

Adaptive Active Learning for Coreference Resolution

Training coreference resolution models require comprehensively labeled d...
research
05/29/2019

Active Learning in the Overparameterized and Interpolating Regime

Overparameterized models that interpolate training data often display su...
research
12/04/2020

Bayesian Active Learning for Wearable Stress and Affect Detection

In the recent past, psychological stress has been increasingly observed ...
research
09/22/2020

Model-Centric and Data-Centric Aspects of Active Learning for Neural Network Models

We study different data-centric and model-centric aspects of active lear...

Please sign up or login with your details

Forgot password? Click here to reset