Presentation

<
GEM3D: GEnerative Medial Abstractions for 3D Shape Synthesis
SessionGenerative 3D Geometry and Editing
DescriptionWe introduce GEM3D — a deep, topology-aware model for generating and reconstructing 3D shapes. Our key ingredient is a neural skeleton-based representation compactly encoding both shape topology and geometry. Experiments show significantly more faithful surface reconstruction and diverse shape generation compared to prior work, especially for structurally complex shapes.
Authors
Dmitrii Petrov
University of Massachusetts Amherst
Pradyumn Goyal
University of Massachusetts Amherst
Vikas Thamizharasan
University of Massachusetts Amherst
Vladimir Kim
Adobe Research
Matheus Gadelha
Adobe Research
Melinos Averkiou
CYENS - Centre of Excellence
University of Cyprus
Siddhartha Chaudhuri
Adobe Research
Evangelos Kalogerakis
University of Massachusetts Amherst
CYENS - Centre of Excellence
Event Type
Technical Paper
TimeWednesday, 31 July 20243:55pm - 4:05pm MDT
LocationMile High 2A
ACM Digital Library Journal Papers' PDFs
Conference Papers' PDFs
Session Time & Location
Sunday, 28 July 20246:00pm - 8:45pm MDTBluebird Ballroom
Wednesday, 31 July 20243:45pm - 5:15pm MDTMile High 2A
Interest Areas
Research & Education
Recordings
Livestreamed
Recorded
Keywords
Geometry
Modeling
Registration Categories
Full Conference
Full Conference Supporter
Virtual Access
Exhibitor Full Conference
Wednesday
Next PresentationNext PresentationCLAY: A Controllable Large-scale Generative Model for Creating High-quality 3D Assets