Masksembles for Uncertainty Estimation

12/15/2020
by   Nikita Durasov, et al.
0

Deep neural networks have amply demonstrated their prowess but estimating the reliability of their predictions remains challenging. Deep Ensembles are widely considered as being one of the best methods for generating uncertainty estimates but are very expensive to train and evaluate. MC-Dropout is another popular alternative, which is less expensive, but also less reliable. Our central intuition is that there is a continuous spectrum of ensemble-like models of which MC-Dropout and Deep Ensembles are extreme examples. The first uses an effectively infinite number of highly correlated models while the second relies on a finite number of independent models. To combine the benefits of both, we introduce Masksembles. Instead of randomly dropping parts of the network as in MC-dropout, Masksemble relies on a fixed number of binary masks, which are parameterized in a way that allows to change correlations between individual models. Namely, by controlling the overlap between the masks and their density one can choose the optimal configuration for the task at hand. This leads to a simple and easy to implement method with performance on par with Ensembles at a fraction of the cost. We experimentally validate Masksembles on two widely used datasets, CIFAR10 and ImageNet.

READ FULL TEXT
research
06/09/2021

Ex uno plures: Splitting One Model into an Ensemble of Subnetworks

Monte Carlo (MC) dropout is a simple and efficient ensembling method tha...
research
07/15/2021

Randomized ReLU Activation for Uncertainty Estimation of Deep Neural Networks

Deep neural networks (DNNs) have successfully learned useful data repres...
research
11/21/2022

ZigZag: Universal Sampling-free Uncertainty Estimation Through Two-Step Inference

Whereas the ability of deep networks to produce useful predictions on ma...
research
06/01/2022

LiDAR-MIMO: Efficient Uncertainty Estimation for LiDAR-based 3D Object Detection

The estimation of uncertainty in robotic vision, such as 3D object detec...
research
10/12/2021

Robust Neural Regression via Uncertainty Learning

Deep neural networks tend to underestimate uncertainty and produce overl...
research
01/02/2019

Can You Trust This Prediction? Auditing Pointwise Reliability After Learning

To use machine learning in high stakes applications (e.g. medicine), we ...
research
01/02/2019

Auditing Pointwise Reliability Subsequent to Training

To use machine learning in high stakes applications (e.g. medicine), we ...

Please sign up or login with your details

Forgot password? Click here to reset