Generative Landmarks

dc.contributor.authorFerman, Daviden_US
dc.contributor.authorBharaj, Gauraven_US
dc.contributor.editorBittner, Jirí and Waldner, Manuelaen_US
dc.date.accessioned2021-04-09T19:18:47Z
dc.date.available2021-04-09T19:18:47Z
dc.date.issued2021
dc.description.abstractWe propose a general purpose approach to detect landmarks with improved temporal consistency, and personalization. Most sparse landmark detection methods rely on laborious, manually labelled landmarks, where inconsistency in annotations over a temporal volume leads to sub-optimal landmark learning. Further, high-quality landmarks with personalization is often hard to achieve. We pose landmark detection as an image translation problem. We capture two sets of unpaired marked (with paint) and unmarked videos. We then use a generative adversarial network and cyclic consistency to predict deformations of landmark templates that simulate markers on unmarked images until these images are indistinguishable from ground-truth marked images. Our novel method does not rely on manually labelled priors, is temporally consistent, and image class agnostic - face, and hand landmarks detection examples are shown.en_US
dc.description.sectionheadersPosters
dc.description.seriesinformationEurographics 2021 - Posters
dc.identifier.doi10.2312/egp.20211029
dc.identifier.isbn978-3-03868-134-2
dc.identifier.issn1017-4656
dc.identifier.pages9-10
dc.identifier.urihttps://doi.org/10.2312/egp.20211029
dc.identifier.urihttps://diglib.eg.org:443/handle/10.2312/egp20211029
dc.publisherThe Eurographics Associationen_US
dc.subjectComputing methodologies
dc.subjectInterest point and salient region detections
dc.subjectTracking
dc.titleGenerative Landmarksen_US
Files
Original bundle
Now showing 1 - 2 of 2
Loading...
Thumbnail Image
Name:
009-010.pdf
Size:
839.02 KB
Format:
Adobe Portable Document Format
Loading...
Thumbnail Image
Name:
poster1009_009-010.pdf
Size:
2.51 MB
Format:
Adobe Portable Document Format