eprintid: 10134583 rev_number: 17 eprint_status: archive userid: 608 dir: disk0/10/13/45/83 datestamp: 2021-09-17 11:46:08 lastmod: 2022-06-15 06:10:54 status_changed: 2021-09-17 11:46:08 type: proceedings_section metadata_visibility: show creators_name: Popescu, SG creators_name: Sharp, DJ creators_name: Cole, JH creators_name: Kamnitsas, K creators_name: Glocker, B title: Distributional Gaussian Process Layers for Outlier Detection in Image Segmentation ispublished: pub divisions: UCL divisions: B04 divisions: C05 divisions: F48 note: This version is the author accepted manuscript. For information on re-use, please refer to the publisher's terms and conditions. abstract: We propose a parameter efficient Bayesian layer for hierarchical convolutional Gaussian Processes that incorporates Gaussian Processes operating in Wasserstein-2 space to reliably propagate uncertainty. This directly replaces convolving Gaussian Processes with a distance-preserving affine operator on distributions. Our experiments on brain tissue-segmentation show that the resulting architecture approaches the performance of well-established deterministic segmentation algorithms (U-Net), which has never been achieved with previous hierarchical Gaussian Processes. Moreover, by applying the same segmentation model to out-of-distribution data (i.e., images with pathology such as brain tumors), we show that our uncertainty estimates result in out-of-distribution detection that outperforms the capabilities of previous Bayesian networks and reconstruction-based approaches that learn normative distributions. date: 2021-01-01 date_type: published publisher: Springer official_url: https://doi.org/10.1007/978-3-030-78191-0_32 oa_status: green full_text_type: other language: eng primo: open primo_central: open_green verified: verified_manual elements_id: 1881977 doi: 10.1007/978-3-030-78191-0_32 isbn_13: 9783030781903 lyricists_name: Cole, James lyricists_id: JCOLE07 actors_name: Cole, James actors_id: JCOLE07 actors_role: owner full_text_status: public series: Lecture Notes in Computer Science publication: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) volume: 12729 place_of_pub: Cham, Switzerland pagerange: 415-427 issn: 1611-3349 book_title: Information Processing in Medical Imaging citation: Popescu, SG; Sharp, DJ; Cole, JH; Kamnitsas, K; Glocker, B; (2021) Distributional Gaussian Process Layers for Outlier Detection in Image Segmentation. In: Information Processing in Medical Imaging. (pp. pp. 415-427). Springer: Cham, Switzerland. Green open access document_url: https://discovery.ucl.ac.uk/id/eprint/10134583/1/Distributional_Gaussian_Process_Layers_for_Outlier_Detection_in_Image_Segmentation____IPMI_version.pdf