Attention enhancement out-of-distribution image detection method based on classifier prediction uncertainty

Abstract

The invention provides an attention enhancement out-of-distribution image detection method based on classifier prediction uncertainty. The uncertainty of the classifier is considered, so that the classification result is more reliable, and the uncertainty is used for weighting the features, so that the classifier pays more attention to the area where the sample is beneficial to classification, and the influence caused by inherent noise of samples in distribution is ignored, so that the effectiveness of the softmax maximum classification probability value in an out-of-distribution sample detection task is greatly improved, and the discrimination of the softmax maximum classification probability value in two types of data is greatly improved. For the out-of-distribution samples, a feature attention graph calculated by utilizing uncertainty can enable a classifier to pay attention to a more wrong region, so that a lower confidence score is caused, and the detection effect is further improved; the method is not excessively limited to training data, misjudgment is not likely to happen to edge samples, and a better detection effect can be obtained.

Description

technical field [0001] The invention relates to the field of out-of-distribution sample detection, and more specifically, to an attention-enhanced out-of-distribution image detection method based on classifier prediction uncertainty. Background technique [0002] Out-of-distribution images are data distributed far from the data used for training, often significantly different from in-distribution data. The method of detecting out-of-distribution data from a large number of in-distribution data is called out-of-distribution sample detection or outlier detection method. In contrast to anomaly detection tasks, out-of-distribution detection is usually applied to labeled data, i.e., considers identifiable categorical features of in-distribution samples of a dataset. Compared with unsupervised anomaly detection methods, the out-of-distribution data coverage in this task is too large to be considered as the same class. [0003] At present, deep learning has become an important me...

AI Technical Summary

Problems solved by technology

Compared with unsupervised anomaly detection methods, the out-of-distribution data coverage in this task is too large to be considered as the same class

[0003] At present, deep learning has become an important means to solve problems, but the problem is that many models can only be successfully applied when the test data and training data are similar, and often cannot be given when encountering some samples that are far from the distribution of training data. reliable results, and may even give a very confident but erroneous prediction

This method is relatively simple and direct, but the effect is not very good compared to other methods

Images