Discriminative out-of-distribution detection for semantic segmentation

Abstract

Most classification and segmentation datasets assume a closed-world scenario in which predictions are expressed as distribution over a predetermined set of visual classes. However, such assumption implies unavoidable and often unnoticeable failures in presence of out-of-distribution (OOD) input. These failures are bound to happen in most real-life applications since current visual ontologies are far from being comprehensive. We propose to address this issue by discriminative detection of OOD pixels in input data. Different from recent approaches, we avoid to bring any decisions by only observing the training dataset of the primary model trained to solve the desired computer vision task. Instead, we train a dedicated OOD model which discriminates the primary training set from a much larger "background" dataset which approximates the variety of the visual world. We perform our experiments on high resolution natural images in a dense prediction setup. We use several road driving datasets as our training distribution, while we approximate the background distribution with the ILSVRC dataset. We evaluate our approach on WildDash test, which is currently the only public test dataset that includes out-of-distribution images. The obtained results show that the proposed approach succeeds to identify out-of-distribution pixels while outperforming previous work by a wide margin.

This paper has been withdrawn from AutoNUE workshop at ECCV 2018 due to ECCV registration being closed

References in corpus (7)

• Rethinking Atrous Convolution for Semantic Image Segmentation
• Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• On Calibration of Modern Neural Networks
• A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
• Understanding Measures of Uncertainty for Adversarial Example Detection
• Robust Semantic Segmentation with Ladder-DenseNet Models

Cited by in corpus (16)

• Energy-based Out-of-distribution Detection
• Scaling Out-of-Distribution Detection for Real-World Settings
• Outlier Exposure with Confidence Control for Out-of-Distribution Detection
• Inspect, Understand, Overcome: A Survey of Practical Methods for AI Safety
• Robust Out-of-distribution Detection for Neural Networks
• Dense outlier detection and open-set recognition based on training with noisy negative images
• ATOM: Robustifying Out-of-distribution Detection Using Outlier Mining
• Synthesize then Compare: Detecting Failures and Anomalies for Semantic Segmentation
• MOS: Towards Scaling Out-of-distribution Detection for Large Semantic Space
• Efficacy of Pixel-Level OOD Detection for Semantic Segmentation
• Pixel-wise Energy-biased Abstention Learning for Anomaly Segmentation on Complex Urban Driving Scenes
• Self-supervised Robust Object Detectors from Partially Labelled Datasets
• Toward Metrics for Differentiating Out-of-Distribution Sets
• Utilizing Network Properties to Detect Erroneous Inputs
• Improving Calibration and Out-of-Distribution Detection in Medical Image Segmentation with Convolutional Neural Networks
• DOODLER: Determining Out-Of-Distribution Likelihood from Encoder Reconstructions