Anomaly Detection

In December 2025, I spoke with Manfred Kloiber on Deutschlandfunk’s Computer und Kommunikation about the safe use of artificial intelligence in railway operations. The discussion covered the conditions under which machine-learning systems can be deployed in safety-critical railway applications, including the challenge of distribution shift, the need for reliable uncertainty quantification, and how learned components interact with established railway safety processes. The interview is in …

Our paper Improving Out-of-Distribution Detection with Markov Logic Networks has been accepted at ICML. In it, we propose a probabilistic extension of Out-of-Distribution Detection with Logical Reasoning, as well as a simple algorithm to mine logical constraints for OOD detection from a dataset. Abstract § Out-of-distribution (OOD) detection is essential for ensuring the reliability of deep learning models operating in open-world scenarios. Current OOD detectors mainly rely on statistical models …

Our paper Out-of-Distribution Detection with Adversarial Outlier Exposure has been accepted at the CVPR workshop for Safe Artificial Intelligence for All Domains (SAIAD). The experiments in the paper were mostly conducted by Thomas Botschen, who is currently a master’s student at our lab. Abstract § Machine learning models typically perform reliably only on inputs drawn from the distribution they were trained on, making Out-of-Distribution (OOD) detection essential for safety-critical …

Our paper, Language Models as Reasoners for Out-of-Distribution Detection, was presented at the Workshop on AI Safety Engineering (WAISE) 2024 and received the best paper award by popular vote. It constitutes an extension of our idea of Out-of-Distribution Detection with Logical Reasoning, where we replaced the prolog-based reasoning component with an LLM. Abstract § Deep neural networks (DNNs) are prone to making wrong predictions with high confidence for data that does not stem from their …

Our paper Out-of-Distribution Detection with Logical Reasoning has been accepted at WACV 2024. Abstract § Machine Learning models often only generalize reliably to samples from the training distribution. Consequentially, detecting when input data is out-of-distribution (OOD) is crucial, especially in safety-critical applications. Current OOD detection methods, however, tend to be domain agnostic and often fail to incorporate valuable prior knowledge about the structure of the training …

My paper Towards Deep Anomaly Detection with Structured Knowledge Representations has been accepted at the Workshop on AI Safety Engineering at SafeComp. Abstract § Machine learning models tend to only make reliable predictions for inputs that are similar to the training data. Consequentially, anomaly detection, which can be used to detect unusual inputs, is critical for ensuring the safety of machine learning agents operating in open environments. In this work, we identify and discuss several …

Our paper On Outlier Exposure with Generative Models has been accepted at the NeurIPS Machine Learning Safety Workshop. Abstract § While Outlier Exposure reliably increases the performance of Out-of-Distribution detectors, it requires a set of available outliers during training. In this paper, we propose Generative Outlier Exposure (GOE), which alleviates the need for available outliers by using generative models to sample synthetic outliers from low-density regions of the data distribution. The …

Our paper Multi-Class Hypersphere Anomaly Detection (MCHAD) has been accepted for presentation at the ICPR 2022. In summary, we propose a new loss function for learning neural networks that are able to detect anomalies in their inputs. Poster for MCHAD (PDF). MCHAD is available via pytorch-ood. You can find example code here. How does it work? § The general idea is that we want a neural network $f_{\theta}: \mathcal{X} \rightarrow \mathcal{Z}$ that maps inputs from the input space to some lower …

Our companion paper, On Challenging Aspects of Reproducibility in Deep Anomaly Detection, has been accepted for presentation at the Fourth Workshop on Reproducible Research in Pattern Recognition (satellite event of ICPR 2022). In it, we discuss aspects of reproducibility for our anomaly detection algorithm MCHAD, as well as anomaly detection with deep neural networks in general. In particular, we discussed the following challenges for reproducibility: Nondeterminism: conducting the same …

Our paper, PyTorch-OOD: A library for Out-of-Distribution Detection based on PyTorch, has been presented at the CVPR 2022 Workshops. You can find the most recent version of the Python source code on GitHub. The library has developed substantially since the original 2022 paper. The current release, documentation, and supported methods are available through the project links below: GitHub · Documentation · PyPI. Abstract § Machine Learning models based on Deep Neural Networks behave unpredictably …