Deep Dirichlet uncertainty for unsupervised out-of-distribution detection of eye fundus photographs in glaucoma screening
This addresses robustness issues in real-world glaucoma screening tools, which is incremental as it builds on existing methods to improve uncertainty estimation without requiring out-of-distribution data.
The paper tackled the problem of overconfident predictions for out-of-distribution cases in glaucoma screening using color fundus photographs, proposing a Dirichlet distribution-based model that achieved the highest average score in the AIROGS challenge as of February 2022.
The development of automatic tools for early glaucoma diagnosis with color fundus photographs can significantly reduce the impact of this disease. However, current state-of-the-art solutions are not robust to real-world scenarios, providing over-confident predictions for out-of-distribution cases. With this in mind, we propose a model based on the Dirichlet distribution that allows to obtain class-wise probabilities together with an uncertainty estimation without exposure to out-of-distribution cases. We demonstrate our approach on the AIROGS challenge. At the start of the final test phase (8 Feb. 2022), our method had the highest average score among all submissions.