Bayesian SegNet: Model Uncertainty in Deep Convolutional Encoder-Decoder Architectures for Scene Understanding

Abstract

We present a deep learning framework for probabilistic pixel-wise semantic segmentation, which we term Bayesian SegNet. Semantic segmentation is an important tool for visual scene understanding and a meaningful measure of uncertainty is essential for decision making. Our contribution is a practical system which is able to predict pixel-wise class labels with a measure of model uncertainty. We achieve this by Monte Carlo sampling with dropout at test time to generate a posterior distribution of pixel class labels. In addition, we show that modelling uncertainty improves segmentation performance by 2-3% across a number of state of the art architectures such as SegNet, FCN and Dilation Network, with no additional parametrisation. We also observe a significant improvement in performance for smaller datasets where modelling uncertainty is more effective. We benchmark Bayesian SegNet on the indoor SUN Scene Understanding and outdoor CamVid driving scenes datasets.

Cited by in corpus (29)

• A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challenges
• Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
• Hybrid LSTM and Encoder-Decoder Architecture for Detection of Image Forgeries
• Modality specific U-Net variants for biomedical image segmentation: A survey
• Uncertainty and Interpretability in Convolutional Neural Networks for Semantic Segmentation of Colorectal Polyps
• Exploiting Epistemic Uncertainty of Anatomy Segmentation for Anomaly Detection in Retinal OCT
• Bayesian-Deep-Learning Estimation of Earthquake Location from Single-Station Observations
• Fusion of Probability Density Functions
• Indoor Scene Understanding in 2.5/3D for Autonomous Agents: A Survey
• SalsaNext: Fast, Uncertainty-aware Semantic Segmentation of LiDAR Point Clouds for Autonomous Driving
• Bayesian convolutional neural network based MRI brain extraction on nonhuman primates
• Multi-Task Learning of Height and Semantics from Aerial Images
• QU-BraTS: MICCAI BraTS 2020 Challenge on Quantifying Uncertainty in Brain Tumor Segmentation - Analysis of Ranking Scores and Benchmarking Results
• The Stixel world: A medium-level representation of traffic scenes
• Fully Convolutional Networks for Chip-wise Defect Detection Employing Photoluminescence Images
• Time-Dynamic Estimates of the Reliability of Deep Semantic Segmentation Networks
• Depth from Monocular Images using a Semi-Parallel Deep Neural Network (SPDNN) Hybrid Architecture
• Self-supervised learning for autonomous vehicles perception: A conciliation between analytical and learning methods
• Deep Network Uncertainty Maps for Indoor Navigation
• Image Processing Methods for Coronal Hole Segmentation, Matching, and Map Classification
• Learning Uncertainty For Safety-Oriented Semantic Segmentation In Autonomous Driving
• Exploring Bayesian Deep Learning for Urgent Instructor Intervention Need in MOOC Forums
• Data Uncertainty Learning in Face Recognition
• Integrating Uncertainty into Neural Network-based Speech Enhancement
• Uncertainty Estimation in Medical Image Localization: Towards Robust Anterior Thalamus Targeting for Deep Brain Stimulation
• Investigating and Improving Latent Density Segmentation Models for Aleatoric Uncertainty Quantification in Medical Imaging
• Bayesian SegNet for Semantic Segmentation with Improved Interpretation of Microstructural Evolution During Irradiation of Materials
• PULASki: Learning inter-rater variability using statistical distances to improve probabilistic segmentation
• Deep Momentum Uncertainty Hashing