17·Research
Academic Papers
Two companion academic papers on applied Gaussian splatting for luxury interiors, one on production-deployment lessons from the benchmark, one (Reflection-Aware 5DGS) on the mirror-and-glass problem. Both complete: four pages each, anonymised for double-blind review, and audited for citation and claim accuracy before being called finished.
[ Ask about this build ]01Discover
Why it exists.
The 3DGS research pipeline turned up a counterintuitive result: in luxury real-estate visualisation, applied Gaussian splatting is bottlenecked by robustness, not quality. The engine with the best average PSNR isn't the one that survives the failure modes a brokerage actually hits.
This paper writes it up: the heatmap as the central artefact, and an argument for reporting failure modes as a first-class benchmark metric, not just averages.
02Define
The brief.
Make the case, in an academic paper, that robustness and failure-mode reporting matter more than leaderboard PSNR for applied 3DGS.
03Develop
What it took
Skills behind it.
Primary discipline plus the support stack.
Skills demonstrated
- ResearchPrimary
Wrote both papers: production-deployment lessons from the benchmark, and Reflection-Aware 5DGS (classified-quadric reflectors for specular interiors).
- 3D & SpatialSecondary
Engine benchmark and scan-difficulty taxonomy from Easy Studio through Extreme Commercial Office.
- AI & Machine LearningSecondary
Novel-view synthesis comparison across diffusion-based and photogrammetric baselines.
- Data PipelinesSupporting
A reproducible harness that computes PSNR, SSIM and LPIPS per run and records per-engine failure modes.
Develop · Notes
The build, in full.
How it came together - kept off this page so the case study stays a read, not a scroll.
Read the full build notes04Deliver
What shipped.
The mirror-and-glass paper already has a convergent result on a real reflective scan. The benchmark paper's eval framework is already used internally to decide which 3DGS engine ships behind Barnes Vantage, per property type.
Research origin
The paper-to-product thread this project sits inside. The full chain, paper through benchmark to shipped, is on the thread.
- 3D Gaussian SplattingSee thread
Turning ordinary photographs of a real place into a 3D scene you can walk through in a browser.
- Mirrors and glass captureSee thread
The open problem Gaussian Splatting still hasn't solved for high-end interiors.
Frames
Selected from the work.
2 frames
01Apartment plan from the scan-to-floor-plan pipeline, 95 m² over 2,190 scan points (grid_v9, April 2026) 02Villa plan from the same pipeline
See also
3D & Spatial
3DGS Research Pipeline
A reproducible harness with 26 Gaussian-splat engines wired into it, scoring which reconstruction survives a real luxury interior.
3D & Spatial
Barnes Vantage
An international portal for viewing real Dubai residences in 3D: walk a real scan yourself, or have a broker drive the same walkthrough live.