Conditioning Latent-Space Clusters for Real-World Anomaly Classification
Not on the map. Why?
Abstract (arXiv)
Anomalies in the domain of autonomous driving are a major hindrance to the large-scale deployment of autonomous vehicles. In this work, we focus on high-resolution camera data from urban scenes that include anomalies of various types and sizes. Based on a Variational Autoencoder, we condition its latent space to classify samples as either normal data or anomalies. In order to emphasize especially small anomalies, we perform experiments where we provide the VAE with a discrepancy map as an additional input, evaluating its impact on the detection performance. Our method separates normal data and anomalies into isolated clusters while still reconstructing high-quality images, leading to meaningful latent representations.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 44 of its 55 references that are in the map. First posted on arXiv 18 Sep 2023.
Builds on
44 of 55 references.
References · 44
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2014Adam: A Method for Stochastic Optimizationcited by 36,960
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
- 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
- 2016Pyramid Scene Parsing Networkcited by 2,472
- 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,362
- 2017Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networkscited by 1,325
- 2020Energy-based Out-of-distribution Detectioncited by 1,296
- 2017High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANscited by 1,176
- 2018Deep Anomaly Detection with Outlier Exposurecited by 939
- 2015Empirical Evaluation of Rectified Activations in Convolutional Networkcited by 515
- 2018GANomaly: Semi-Supervised Anomaly Detection via Adversarial Trainingcited by 409
32 more
- 2022VOS: Learning What You Don't Know by Virtual Outlier Synthesiscited by 300
- 2020Learning Memory-guided Normality for Anomaly Detectioncited by 278
- 2016Towards K-means-friendly Spaces: Simultaneous Deep Learning and Clusteringcited by 237
- 2016Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoderscited by 219
- 2021SegmentMeIfYouCan: A Benchmark for Anomaly Segmentationcited by 150
- 2018Latent Space Autoregression for Novelty Detectioncited by 149
- 2019Detecting the Unexpected via Image Resynthesiscited by 137
- 2019The Fishyscapes Benchmark: Measuring Blind Spots in Semantic Segmentationcited by 122
- 2016Lost and Found: Detecting Small Road Hazards for Self-Driving Vehiclescited by 117
- 2019Towards Visually Explaining Variational Autoencoderscited by 113
- 2021Pixel-wise Anomaly Detection in Complex Driving Scenescited by 109
- 2022Anomaly Detection in Autonomous Driving: A Surveycited by 87
- 2018Anomaly Detection using One-Class Neural Networkscited by 82
- 2021Deep Metric Learning for Open World Semantic Segmentationcited by 51
- 2019Towards Corner Case Detection for Autonomous Drivingcited by 39
- 2020Out-of-Distribution Detection for Automotive Perceptioncited by 37
- 2021Inspect, Understand, Overcome: A Survey of Practical Methods for AI Safetycited by 35
- 2021Description of Corner Cases in Automated Driving: Goals and Challengescited by 35
- 2021An Application-Driven Conceptualization of Corner Cases for Perception in Highly Automated Drivingcited by 33
- 2021Corner Cases for Visual Perception in Automated Driving: Some Guidance on Detection Approachescited by 27
- 2019Anomaly Detection with Adversarial Dual Autoencoderscited by 16
- 2023Perception Datasets for Anomaly Detection in Autonomous Driving: A Surveycited by 16
- 2020Out-of-Distribution Detection in Multi-Label Datasets using Latent Space of β-VAEcited by 15
- 2022Ad-datasets: a meta-collection of data sets for autonomous drivingcited by 11
- 2020Interpreting Rate-Distortion of Variational Autoencoder and Using Model Uncertainty for Anomaly Detectioncited by 10
- 2019Latent space conditioning for improved classification and anomaly detectioncited by 8
- 2020Deep Representation Learning and Clustering of Traffic Scenarioscited by 7
- 2019Implicit Discriminator in Variational Autoencodercited by 5
- 2021Compressing Sensor Data for Remote Assistance of Autonomous Vehicles using Deep Generative Modelscited by 5
- 2021Novelty Detection and Analysis of Traffic Scenario Infrastructures in the Latent Space of a Vision Transformer-Based Triplet Autoencodercited by 4
- 2023Impact, Attention, Influence: Early Assessment of Autonomous Driving Datasetscited by 4
- 2021What is Wrong with One-Class Anomaly Detection?cited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.