Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386249571> ?p ?o ?g. }
Showing items 1 to 87 of
87
with 100 items per page.
- W4386249571 abstract "Safety measures need to be systemically investigated to what extent they evaluate the intended performance of Deep Neural Networks (DNNs) for critical applications. Due to a lack of verification methods for high-dimensional DNNs, a trade-off is needed between accepted performance and handling of out-of-distribution (OOD) samples.This work evaluates rejecting outputs from semantic segmentation DNNs by applying a Mahalanobis distance (MD) based on the most probable class-conditional Gaussian distribution for the predicted class as an OOD score. The evaluation follows three DNNs trained on the Cityscapes dataset and tested on four automotive datasets and finds that classification risk can drastically be reduced at the cost of pixel coverage, even when applied on unseen datasets. The applicability of our findings will support legitimizing safety measures and motivate their usage when arguing for safe usage of DNNs in automotive perception." @default.
- W4386249571 created "2023-08-30" @default.
- W4386249571 creator A5004374106 @default.
- W4386249571 creator A5011682167 @default.
- W4386249571 creator A5015885229 @default.
- W4386249571 creator A5023822114 @default.
- W4386249571 date "2023-07-01" @default.
- W4386249571 modified "2023-09-27" @default.
- W4386249571 title "Evaluation of Out-of-Distribution Detection Performance on Autonomous Driving Datasets" @default.
- W4386249571 cites W2015887370 @default.
- W4386249571 cites W2149699531 @default.
- W4386249571 cites W2340897893 @default.
- W4386249571 cites W2560023338 @default.
- W4386249571 cites W2888307014 @default.
- W4386249571 cites W2919275752 @default.
- W4386249571 cites W2962752734 @default.
- W4386249571 cites W2963149653 @default.
- W4386249571 cites W3035564946 @default.
- W4386249571 cites W3036939397 @default.
- W4386249571 cites W3084917217 @default.
- W4386249571 cites W3120251014 @default.
- W4386249571 cites W3160165418 @default.
- W4386249571 cites W3176640650 @default.
- W4386249571 cites W3203597819 @default.
- W4386249571 cites W3211597178 @default.
- W4386249571 cites W3216928200 @default.
- W4386249571 cites W4211087648 @default.
- W4386249571 cites W4292787455 @default.
- W4386249571 cites W4313547544 @default.
- W4386249571 cites W4362513284 @default.
- W4386249571 doi "https://doi.org/10.1109/aitest58265.2023.00021" @default.
- W4386249571 hasPublicationYear "2023" @default.
- W4386249571 type Work @default.
- W4386249571 citedByCount "0" @default.
- W4386249571 crossrefType "proceedings-article" @default.
- W4386249571 hasAuthorship W4386249571A5004374106 @default.
- W4386249571 hasAuthorship W4386249571A5011682167 @default.
- W4386249571 hasAuthorship W4386249571A5015885229 @default.
- W4386249571 hasAuthorship W4386249571A5023822114 @default.
- W4386249571 hasConcept C119857082 @default.
- W4386249571 hasConcept C121332964 @default.
- W4386249571 hasConcept C124101348 @default.
- W4386249571 hasConcept C127413603 @default.
- W4386249571 hasConcept C146978453 @default.
- W4386249571 hasConcept C153180895 @default.
- W4386249571 hasConcept C154945302 @default.
- W4386249571 hasConcept C163716315 @default.
- W4386249571 hasConcept C1921717 @default.
- W4386249571 hasConcept C2777212361 @default.
- W4386249571 hasConcept C2984842247 @default.
- W4386249571 hasConcept C41008148 @default.
- W4386249571 hasConcept C50644808 @default.
- W4386249571 hasConcept C526921623 @default.
- W4386249571 hasConcept C62520636 @default.
- W4386249571 hasConcept C89600930 @default.
- W4386249571 hasConceptScore W4386249571C119857082 @default.
- W4386249571 hasConceptScore W4386249571C121332964 @default.
- W4386249571 hasConceptScore W4386249571C124101348 @default.
- W4386249571 hasConceptScore W4386249571C127413603 @default.
- W4386249571 hasConceptScore W4386249571C146978453 @default.
- W4386249571 hasConceptScore W4386249571C153180895 @default.
- W4386249571 hasConceptScore W4386249571C154945302 @default.
- W4386249571 hasConceptScore W4386249571C163716315 @default.
- W4386249571 hasConceptScore W4386249571C1921717 @default.
- W4386249571 hasConceptScore W4386249571C2777212361 @default.
- W4386249571 hasConceptScore W4386249571C2984842247 @default.
- W4386249571 hasConceptScore W4386249571C41008148 @default.
- W4386249571 hasConceptScore W4386249571C50644808 @default.
- W4386249571 hasConceptScore W4386249571C526921623 @default.
- W4386249571 hasConceptScore W4386249571C62520636 @default.
- W4386249571 hasConceptScore W4386249571C89600930 @default.
- W4386249571 hasLocation W43862495711 @default.
- W4386249571 hasOpenAccess W4386249571 @default.
- W4386249571 hasPrimaryLocation W43862495711 @default.
- W4386249571 hasRelatedWork W1991269640 @default.
- W4386249571 hasRelatedWork W2033000528 @default.
- W4386249571 hasRelatedWork W2097836861 @default.
- W4386249571 hasRelatedWork W2902494752 @default.
- W4386249571 hasRelatedWork W2961085424 @default.
- W4386249571 hasRelatedWork W3044078048 @default.
- W4386249571 hasRelatedWork W3161321444 @default.
- W4386249571 hasRelatedWork W4246585671 @default.
- W4386249571 hasRelatedWork W4306674287 @default.
- W4386249571 hasRelatedWork W4224009465 @default.
- W4386249571 isParatext "false" @default.
- W4386249571 isRetracted "false" @default.
- W4386249571 workType "article" @default.