{"id":43812,"date":"2026-11-13T10:59:33","date_gmt":"2026-11-13T10:59:33","guid":{"rendered":"https:\/\/lamarr-institute.org\/publication\/unrealmvs-a-generator-for-synthetic-multi-fisheye-omnidirectional-depth-data\/"},"modified":"2026-11-13T10:59:33","modified_gmt":"2026-11-13T10:59:33","slug":"unrealmvs-a-generator-for-synthetic-multi-fisheye-omnidirectional-depth-data","status":"publish","type":"publication","link":"https:\/\/lamarr-institute.org\/de\/publication\/unrealmvs-a-generator-for-synthetic-multi-fisheye-omnidirectional-depth-data\/","title":{"rendered":"{UnrealMVS}: A Generator for Synthetic Multi-Fisheye Omnidirectional Depth Data"},"content":{"rendered":"<p>Wide-angle and omnidirectional camera systems are increasingly used in robotics and autonomous vehicles, yet the development of depth estimation methods for such sensors is fundamentally constrained by the scarcity of calibrated multifisheye datasets with dense depth ground truth. This paper presents a fully parametrizable dataset-generation pipeline for Unreal Engine 5 ({UE}5) that produces synchronized fisheye and equirectangular {RGB} images with per-frame metric depth, intrinsic camera parameters and relative extrinsics. Since {UE}5 provides neither native fisheye sensors nor reliable equirectangular depth rendering, we implement a cubemap-based approach supporting configurable projection models including Kannala Brandt ({KB}4) and Enhanced Unified Camera Model ({EUCM}). We release both the pipeline and a collection of synthetic datasets generated across diverse environments, and validate the data by retraining an omnidirectional Multi-View Stereo ({MVS}) network, demonstrating improved generalization when combining our data with existing datasets.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Wide-angle and omnidirectional camera systems are increasingly used in robotics and autonomous vehicles, yet the development of depth estimation methods for such sensors is fundamentally constrained by the scarcity of calibrated multifisheye datasets with dense depth ground truth. This paper presents a fully parametrizable dataset-generation pipeline for Unreal Engine 5 ({UE}5) that produces synchronized fisheye and equirectangular {RGB} images with per-frame metric depth, intrinsic camera parameters and relative extrinsics. Since [&hellip;]<\/p>\n","protected":false},"author":14,"featured_media":0,"template":"","meta":{"_acf_changed":false,"footnotes":""},"publication-type":[32],"publication-year":[1680],"class_list":["post-43812","publication","type-publication","status-publish","hentry","publication-type-inproceedings","publication-year-1680"],"acf":[],"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication\/43812","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication"}],"about":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/types\/publication"}],"author":[{"embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/users\/14"}],"version-history":[{"count":0,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication\/43812\/revisions"}],"wp:attachment":[{"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/media?parent=43812"}],"wp:term":[{"taxonomy":"publication-type","embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication-type?post=43812"},{"taxonomy":"publication-year","embeddable":true,"href":"https:\/\/lamarr-institute.org\/de\/wp-json\/wp\/v2\/publication-year?post=43812"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}