← Research Timeline Aditya Jain / Apple Maps · 3D Reconstruction
Oct 2025
Topic 18 Oct 2025 Paper Survey · Field Gap Analysis

Research Survey —
CAPRI-Net, BrepGen, HoLa, SparC3D, TRELLIS.

Four-paper survey of state-of-the-art 3-D reconstruction methods, each studied from paper read → GitHub clone → inference attempt. The output: identification of the field gap that the thesis line targets — no published method combines image-to-3-D with CAD-parametric (procedural) control. Either you get neural reconstruction (high fidelity, no parameters) or you get procedural generation (low fidelity, full parameters). The thesis line's goal is to bridge that gap.

00 — Motivation

Find the gap that justifies the thesis line.

Before committing to the thesis-line architectural decisions (procedural + neural hybrid, triplane intermediate, MambaFlow3D- class generators), the obvious question is whether someone has already shipped what the thesis is trying to build. The October 2025 survey work was the field-gap-finding exercise. Four representative recent papers were each read end-to-end, cloned, and inference-attempted on local hardware (Intel iMac, RTX 3060 via Vast.ai, or Colab T4 depending on the paper's compute requirements).

The finding that justified the thesis-line scope: the field bifurcates into neural reconstruction methods (SparC3D, TRELLIS, HoLa) that produce high-fidelity 3-D from images but have no parametric / procedural interface; and procedural CAD generation methods (CAPRI-Net, BrepGen) that produce fully-parametric output but from sparse / latent input rather than from images. The thesis line targets the union — image-to- 3-D with procedural / parametric output — which no paper at the time of survey solved end-to-end.

01 — Paper-by-Paper
PaperApproachInputOutputParametric?
CAPRI-NetCSG primitive composition via learned programPoint cloudCSG program (primitives + operations)Yes
BrepGenDiffusion on boundary-representation graphLatent codeB-rep CAD modelYes
HoLaHierarchical learned B-rep generationLatent codeB-rep CAD model with topologyYes
SparC3DSparse-cube transformer over voxel tokensSingle imageSparse-voxel 3-D meshNo
TRELLISStructured latent + dual decoderSingle imageSparse-voxel + Gaussian splat outputNo
The Field Gap

Image-to-3-D ∩ Parametric = ∅.
No paper does both. That's the thesis-line opportunity.

SparC3D and TRELLIS demonstrate state-of-the-art neural image- to-3-D with no procedural interface. CAPRI-Net, BrepGen, and HoLa demonstrate state-of-the-art procedural / parametric CAD generation with no image input. The intersection — the capability that the Apple Maps procedural-modelling team needs most — is empty. The thesis line targets that intersection explicitly: image → triplane → procedural DSL (PGN-class) / primitive program (SculptNet-class).

02 — Local-Inference Attempts
PaperLocal hardware triedOutcome
CAPRI-NetIntel iMac, then Jupyter notebook with GPUSuccessful inference after dependency-pin fixes
BrepGenGoogle Colab T4Successful after CUDA-version compatibility fix
HoLaiMac (CPU)Inference too slow to be useful; queued for RTX 3060 retry
SparC3DHugging Face Space (cloud)Works; analysed via the HF demo rather than local install
TRELLISMicrosoft research release · A100-class requiredRead-only — local hardware insufficient

Interactive Demo · Live

The field gap as a 2-D scatter plot. Each paper is a dot; X-axis is "image input capability", Y-axis is "parametric output". The empty top-right quadrant is the thesis-line target.

01 — Field map PICK A PAPER
02 — Paper details on hover
Click a dot on the field map to read the paper's approach, input, and parametric capability.
03 — Thesis-line position top-right quadrant
The thesis-line target sits in the empty top-right quadrant of the field map: image input + parametric output.

PGN (Topic 40) is the first concrete attempt at this — polyline + attribute input → executable DSL output. SculptNet (Topic 36) is the second — image input → primitive-program output. Both populate the field-gap quadrant.

Full Technical Paper

White paper · five-paper field-gap survey · image-input × parametric-output quadrant · thesis-line scope decision

Read Paper →
Related Thesis Chapters
PGN
First thesis-line entry into the field-gap quadrant (image / polyline → parametric DSL).
SculptNet
Second entry — image → 5-primitive parametric program; the most direct response to the survey's gap finding.
G-Splats vs VDB
Follow-up survey work focusing on the neural-rendering side of the field, complementing this CAD-side survey.
Appendix — Raw Materials
Transcripts & Source References
████████████████████████████████████████████████

██████████████████████████████████████
█████████ · ████ · █████████████████████
████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████
Restricted Access