That Label's Got Style: Handling Label Style Bias for Uncertain Image Segmentation
arXiv:2303.15850
Abstract
Segmentation uncertainty models predict a distribution over plausible segmentations for a given input, which they learn from the annotator variation in the training set. However, in practice these annotations can differ systematically in the way they are generated, for example through the use of different labeling tools. This results in datasets that contain both data variability and differing label styles. In this paper, we demonstrate that applying state-of-the-art segmentation uncertainty models on such datasets can lead to model bias caused by the different label styles. We present an updated modelling objective conditioning on labeling style for aleatoric uncertainty estimation, and modify two state-of-the-art-architectures for segmentation uncertainty accordingly. We show with extensive experiments that this method reduces label style bias, while improving segmentation performance, increasing the applicability of segmentation uncertainty models in the wild. We curate two datasets, with annotations in different label styles, which we will make publicly available along with our code upon publication.
References in corpus (10)
- Conditional Generative Adversarial Nets
- Efficient Multi-Scale 3D CNN with Fully Connected CRF for Accurate Brain Lesion Segmentation
- What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
- Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning
- Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inference
- BCN20000: Dermoscopic Lesions in the Wild
- A Survey of Uncertainty in Deep Neural Networks
- Stochastic Segmentation Networks: Modelling Spatially Correlated Aleatoric Uncertainty
- Dropout as a Bayesian Approximation: Appendix
- Is MC Dropout Bayesian?