GLINT (CVPR 2026 Oral): Solving Transparency in 3D Gaussian Splatting via Radiance Decomposition
rsasaki0109 · x · 2026-08-13
Selected as a CVPR 2026 Oral (Award Candidate), the GLINT framework by KAIST and NAVER LABS tackles the fundamental inability of 3D Gaussian Splatting (3DGS) to render transparent materials like glass.
Key Methodology:
- Explicitly decomposes outgoing radiance into interface, transmission, and reflection components.
- Leverages geometry and material priors from a pre-trained video relighting model.
- Bootstraps transparency localization using geometry-separation cues induced by the decomposition.
Extensive experiments show that GLINT significantly outperforms prior methods in reconstructing complex transparent scenes.
More from Multimodal
- XGRIDS builds world's largest 3DGS town: 23.5 km² and 6.2 billion Gaussian Splats — bilawalsidhu · 2026-10-16
- IMMERSION: AI film compilation links 12 shorts with one continuous camera move — AI_Andrew · 2026-10-16
- A song with vocals, lyrics and 3D head animation created entirely by Opus 5.5 — repligate · 2026-10-16
- Alibaba open-sources CommerceGen-Studio for commercial-grade e-commerce images — alibabagroup · 2026-10-16
- EVA-Cap: Fudan and OPPO align video captioning with atomic audiovisual events via GRPO — Fudan-University · 2026-10-16
- SCOPED: skill-decoupled expert distillation beats mixed-capability RL across 8 benchmarks — Yuqi Xiong · 2026-10-16