This article is a Preprint
Preprints are preliminary research reports that have not been certified by peer review. They should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Preprints posted online allow authors to receive rapid feedback and the entire scientific community can appraise the work for themselves and respond appropriately. Those comments are posted alongside the preprints for anyone to read them and serve as a post publication assessment.
An Algorithm for Out-Of-Distribution Attack to Neural Network Encoder (preprint)
arxiv; 2020.
Preprint
in English
| PREPRINT-ARXIV | ID: ppzbmed-2009.08016v3
ABSTRACT
Deep neural network (DNN), especially convolutional neural network, has achieved superior performance on image classification tasks. However, such performance is only guaranteed if the input to a trained model is similar to the training samples, i.e., the input follows the probability distribution of the training set. Out-Of-Distribution (OOD) samples do not follow the distribution of training set, and therefore the predicted class labels on OOD samples become meaningless. Classification-based methods have been proposed for OOD detection; however, in this study we show that this type of method is theoretically ineffective and practically breakable because of dimensionality reduction in the model. We also show that Glow likelihood-based OOD detection is ineffective as well. Our analysis is demonstrated on five open datasets, including a COVID-19 CT dataset. At last, we present a simple theoretical solution with guaranteed performance for OOD detection.
Full text:
Available
Collection:
Preprints
Database:
PREPRINT-ARXIV
Main subject:
COVID-19
Language:
English
Year:
2020
Document Type:
Preprint
Similar
MEDLINE
...
LILACS
LIS