Matches in SemOpenAlex for { <https://semopenalex.org/work/W4307941493> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W4307941493 endingPage "8331" @default.
- W4307941493 startingPage "8331" @default.
- W4307941493 abstract "Solid developments have been seen in deep-learning-based pose estimation, but few works have explored performance in dense crowds, such as a classroom scene; furthermore, no specific knowledge is considered in the design of image augmentation for pose estimation. A masked autoencoder was shown to have a non-negligible capability in image reconstruction, where the masking mechanism that randomly drops patches forces the model to build unknown pixels from known pixels. Inspired by this self-supervised learning method, where the restoration of the feature loss induced by the mask is consistent with tackling the occlusion problem in classroom scenarios, we discovered that the transfer performance of the pre-trained weights could be used as a model-based augmentation to overcome the intractable occlusion in classroom pose estimation. In this study, we proposed a top-down pose estimation method that utilized the natural reconstruction capability of missing information of the MAE as an effective occluded image augmentation in a pose estimation task. The difference with the original MAE was that instead of using a 75% random mask ratio, we regarded the keypoint distribution probabilistic heatmap as a reference for masking, which we named Pose Mask. To test the performance of our method in heavily occluded classroom scenes, we collected a new dataset for pose estimation in classroom scenes named Class Pose and conducted many experiments, the results of which showed promising performance." @default.
- W4307941493 created "2022-11-06" @default.
- W4307941493 creator A5007457930 @default.
- W4307941493 creator A5011371569 @default.
- W4307941493 creator A5015102287 @default.
- W4307941493 creator A5016199726 @default.
- W4307941493 creator A5041319629 @default.
- W4307941493 date "2022-10-30" @default.
- W4307941493 modified "2023-10-01" @default.
- W4307941493 title "Pose Mask: A Model-Based Augmentation Method for 2D Pose Estimation in Classroom Scenes Using Surveillance Images" @default.
- W4307941493 cites W2928165649 @default.
- W4307941493 cites W2998625377 @default.
- W4307941493 cites W3096831136 @default.
- W4307941493 cites W3133964188 @default.
- W4307941493 cites W3135266094 @default.
- W4307941493 cites W3191120830 @default.
- W4307941493 doi "https://doi.org/10.3390/s22218331" @default.
- W4307941493 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/36366027" @default.
- W4307941493 hasPublicationYear "2022" @default.
- W4307941493 type Work @default.
- W4307941493 citedByCount "0" @default.
- W4307941493 crossrefType "journal-article" @default.
- W4307941493 hasAuthorship W4307941493A5007457930 @default.
- W4307941493 hasAuthorship W4307941493A5011371569 @default.
- W4307941493 hasAuthorship W4307941493A5015102287 @default.
- W4307941493 hasAuthorship W4307941493A5016199726 @default.
- W4307941493 hasAuthorship W4307941493A5041319629 @default.
- W4307941493 hasBestOaLocation W43079414931 @default.
- W4307941493 hasConcept C101738243 @default.
- W4307941493 hasConcept C108583219 @default.
- W4307941493 hasConcept C115961682 @default.
- W4307941493 hasConcept C138885662 @default.
- W4307941493 hasConcept C142362112 @default.
- W4307941493 hasConcept C153180895 @default.
- W4307941493 hasConcept C153349607 @default.
- W4307941493 hasConcept C154945302 @default.
- W4307941493 hasConcept C160633673 @default.
- W4307941493 hasConcept C2776401178 @default.
- W4307941493 hasConcept C2777402240 @default.
- W4307941493 hasConcept C31972630 @default.
- W4307941493 hasConcept C36613465 @default.
- W4307941493 hasConcept C41008148 @default.
- W4307941493 hasConcept C41895202 @default.
- W4307941493 hasConcept C52102323 @default.
- W4307941493 hasConceptScore W4307941493C101738243 @default.
- W4307941493 hasConceptScore W4307941493C108583219 @default.
- W4307941493 hasConceptScore W4307941493C115961682 @default.
- W4307941493 hasConceptScore W4307941493C138885662 @default.
- W4307941493 hasConceptScore W4307941493C142362112 @default.
- W4307941493 hasConceptScore W4307941493C153180895 @default.
- W4307941493 hasConceptScore W4307941493C153349607 @default.
- W4307941493 hasConceptScore W4307941493C154945302 @default.
- W4307941493 hasConceptScore W4307941493C160633673 @default.
- W4307941493 hasConceptScore W4307941493C2776401178 @default.
- W4307941493 hasConceptScore W4307941493C2777402240 @default.
- W4307941493 hasConceptScore W4307941493C31972630 @default.
- W4307941493 hasConceptScore W4307941493C36613465 @default.
- W4307941493 hasConceptScore W4307941493C41008148 @default.
- W4307941493 hasConceptScore W4307941493C41895202 @default.
- W4307941493 hasConceptScore W4307941493C52102323 @default.
- W4307941493 hasIssue "21" @default.
- W4307941493 hasLocation W43079414931 @default.
- W4307941493 hasLocation W43079414932 @default.
- W4307941493 hasLocation W43079414933 @default.
- W4307941493 hasLocation W43079414934 @default.
- W4307941493 hasOpenAccess W4307941493 @default.
- W4307941493 hasPrimaryLocation W43079414931 @default.
- W4307941493 hasRelatedWork W1968716783 @default.
- W4307941493 hasRelatedWork W2004095265 @default.
- W4307941493 hasRelatedWork W2064877078 @default.
- W4307941493 hasRelatedWork W2085100591 @default.
- W4307941493 hasRelatedWork W2128635338 @default.
- W4307941493 hasRelatedWork W2154613448 @default.
- W4307941493 hasRelatedWork W2286009621 @default.
- W4307941493 hasRelatedWork W2567319754 @default.
- W4307941493 hasRelatedWork W3171518929 @default.
- W4307941493 hasRelatedWork W4306804934 @default.
- W4307941493 hasVolume "22" @default.
- W4307941493 isParatext "false" @default.
- W4307941493 isRetracted "false" @default.
- W4307941493 workType "article" @default.