3D-SSGAN: Lifting 2D Semantics for 3D-Aware Compositional Portrait Synthesis

Simple item page
dc.contributor.authorLiu, Ruiqien_US
dc.contributor.authorZheng, Pengen_US
dc.contributor.authorWang, Yeen_US
dc.contributor.authorMa, Ruien_US
dc.contributor.editorChen, Renjieen_US
dc.contributor.editorRitschel, Tobiasen_US
dc.contributor.editorWhiting, Emilyen_US
dc.date.accessioned2024-10-13T18:05:39Z
dc.date.available2024-10-13T18:05:39Z
dc.date.issued2024
dc.description.abstractExisting 3D-aware portrait synthesis methods can generate impressive high-quality images while preserving strong 3D consistency. However, most of them cannot support the fine-grained part-level control over synthesized images. Conversely, some GAN-based 2D portrait synthesis methods can achieve clear disentanglement of facial regions, but they cannot preserve view consistency due to a lack of 3D modeling abilities. To address these issues, we propose 3D-SSGAN, a novel framework for 3D-aware compositional portrait image synthesis. First, a simple yet effective depth-guided 2D-to-3D lifting module maps the generated 2D part features and semantics to 3D. Then, a volume renderer with a novel 3D-aware semantic mask renderer is utilized to produce the composed face features and corresponding masks. The whole framework is trained end-to-end by discriminating between real and synthesized 2D images and their semantic masks. Quantitative and qualitative evaluations demonstrate the superiority of 3D-SSGAN in controllable part-level synthesis while preserving 3D view consistency.en_US
dc.description.sectionheadersAdvanced 3D Synthesis and Stylization
dc.description.seriesinformationPacific Graphics Conference Papers and Posters
dc.identifier.doi10.2312/pg.20241319
dc.identifier.isbn978-3-03868-250-9
dc.identifier.pages11 pages
dc.identifier.urihttps://doi.org/10.2312/pg.20241319
dc.identifier.urihttps://diglib.eg.org/handle/10.2312/pg20241319
dc.publisherThe Eurographics Associationen_US
dc.rightsAttribution 4.0 International License
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.subjectCCS Concepts: Computing methodologies → Image manipulation
dc.subjectComputing methodologies → Image manipulation
dc.title3D-SSGAN: Lifting 2D Semantics for 3D-Aware Compositional Portrait Synthesisen_US
Files
Original bundle
Now showing 1 - 3 of 3
Thumbnail Image
Name:
pg20241319.pdf
Size:
7.93 MB
Format:
Adobe Portable Document Format
Download
No Thumbnail Available
Name:
paper1207_mm.mp4
Size:
7.9 MB
Format:
Video MP4
Download
Thumbnail Image
Name:
paper1207_mm.pdf
Size:
6.24 MB
Format:
Adobe Portable Document Format
Download
Collections
PG2024 Conference Papers and Posters