no code implementations • 8 Jan 2024 • Jonas Dippel, Barbara Feulner, Tobias Winterhoff, Simon Schallenberg, Gabriel Dernbach, Andreas Kunft, Stephan Tietz, Philipp Jurmeister, David Horst, Lukas Ruff, Klaus-Robert Müller, Frederick Klauschen, Maximilian Alber
Histopathology plays a central role in clinical medicine and biomedical research.
Motivated by the medical application, we demonstrate for general segmentation tasks that including additional patches with solely weak complementary labels during model training can significantly improve the predictive performance and robustness of a model.
We find that standard classifiers and semi-supervised one-class methods trained to discern between normal samples and relatively few random natural images are able to outperform the current state of the art on an established AD benchmark with ImageNet.
Ranked #1 on Anomaly Detection on One-class CIFAR-10 (using extra training data)
Humans find structure in natural phenomena by absorbing stimuli from multiple input sources such as vision, text, and speech.
Deep anomaly detection is a difficult task since, in high dimensions, it is hard to completely characterize a notion of "differentness" when given only examples of normality.
Deep learning approaches to anomaly detection have recently improved the state of the art in detection performance on complex datasets such as large collections of images or text.
Deep one-class classification variants for anomaly detection learn a mapping that concentrates nominal samples in feature space causing anomalies to be mapped away.
Ranked #5 on Anomaly Detection on One-class ImageNet-30 (using extra training data)
The 'Clever Hans' effect occurs when the learned model produces correct predictions based on the 'wrong' features.
Though anomaly detection (AD) can be viewed as a classification problem (nominal vs. anomalous) it is usually treated in an unsupervised manner since one typically does not have access to, or it is infeasible to utilize, a dataset that sufficiently characterizes what it means to be "anomalous."
However, deep SVDD suffers from hypersphere collapse -- also known as mode collapse, if the architecture of the model does not comply with certain architectural constraints, e. g. the removal of bias terms.
There exist few text-specific methods for unsupervised anomaly detection, and for those that do exist, none utilize pre-trained models for distributed vector representations of words.
Cluster predictions of the obtained networks can then be quickly and accurately attributed to the input features.
Deep approaches to anomaly detection have recently shown promising results over shallow methods on large and complex datasets.
Despite the great advances made by deep learning in many machine learning problems, there is a relative dearth of deep learning approaches for anomaly detection.
Ranked #32 on Anomaly Detection on One-class CIFAR-10
Many anomaly detection methods exist that perform well on low-dimensional problems however there is a notable lack of effective methods for high-dimensional spaces, such as images.