Drive-Through 3D Vehicle Exterior Reconstruction via Dynamic-Scene SfM and Distortion-Aware Gaussian Splatting

Drive-Through 3D Vehicle Exterior Reconstruction via Dynamic-Scene SfM and Distortion-Aware Gaussian Splatting | Signal Canvas | ScienceToStartup

Page Freshness

Signal Canvas proof surface

Canonical route: /signal-canvas/drive-through-3d-vehicle-exterior-reconstruction-via-dynamic-scene-sfm-and-distortion-aware-gaussian-splatting

stale

Proof freshness: stale
Proof status: unverified
Display score: 7/10
Last proof check: 2026-03-30
Score updated: 2026-04-02
Score fresh until: 2026-05-02
References: 24
Source count: 3
Coverage: 50%

This page is showing the last landed evidence receipt and score bundle because the latest proof data is outside the freshness window.

Agent Handoff

Canonical ID drive-through-3d-vehicle-exterior-reconstruction-via-dynamic-scene-sfm-and-distortion-aware-gaussian-splatting | Route /signal-canvas/drive-through-3d-vehicle-exterior-reconstruction-via-dynamic-scene-sfm-and-distortion-aware-gaussian-splatting

REST example

curl https://sciencetostartup.com/api/v1/agent-handoff/signal-canvas/drive-through-3d-vehicle-exterior-reconstruction-via-dynamic-scene-sfm-and-distortion-aware-gaussian-splatting

MCP example

{
  "tool": "search_signal_canvas",
  "arguments": {
    "mode": "paper",
    "paper_ref": "drive-through-3d-vehicle-exterior-reconstruction-via-dynamic-scene-sfm-and-distortion-aware-gaussian-splatting",
    "query_text": "Summarize Drive-Through 3D Vehicle Exterior Reconstruction via Dynamic-Scene SfM and Distortion-Aware Gaussian Splatting"
  }
}

source_context

{
  "surface": "signal_canvas",
  "mode": "paper",
  "query": "Drive-Through 3D Vehicle Exterior Reconstruction via Dynamic-Scene SfM and Distortion-Aware Gaussian Splatting",
  "normalized_query": "2603.26638",
  "route": "/signal-canvas/drive-through-3d-vehicle-exterior-reconstruction-via-dynamic-scene-sfm-and-distortion-aware-gaussian-splatting",
  "paper_ref": "drive-through-3d-vehicle-exterior-reconstruction-via-dynamic-scene-sfm-and-distortion-aware-gaussian-splatting",
  "topic_slug": null,
  "benchmark_ref": null,
  "dataset_ref": null
}

Evidence Receipt

Route status: building

Claims: 12

References: 24

Proof: Verification pending

Freshness state: computing

Source paper: Drive-Through 3D Vehicle Exterior Reconstruction via Dynamic-Scene SfM and Distortion-Aware Gaussian Splatting

PDF: https://arxiv.org/pdf/2603.26638v1

Source count: 3

Coverage: 50%

Last proof check: 2026-03-30T21:51:41.406Z

Signal Canvas receipt window

Watch and verify: Drive-Through 3D Vehicle Exterior Reconstruction via Dynamic-Scene SfM and Distortion-Aware Gaussian Splatting

/buildability/drive-through-3d-vehicle-exterior-reconstruction-via-dynamic-scene-sfm-and-distortion-aware-gaussian-splatting

Watchwatch

Subject: Drive-Through 3D Vehicle Exterior Reconstruction via Dynamic-Scene SfM and Distortion-Aware Gaussian Splatting

Verdict

Watch

Verdict is Watch because viability or proof quality is intermediate and should be re-evaluated before execution.

Preparing verified analysis

GitHub Code Pulse

No public code linked for this paper yet.

Claim map

Strong 12Mixed 0Weak 0

Evidencepartial
representing a 3.85 dB improvement over standard 3D-GS
Implicationpartial
This claim is explicitly stated in the abstract with a specific quantitative improvement.
Verificationpartial
partial
Evidencepartial
We propose an end-to-end pipeline utilizing a two-pillar camera rig.
Implicationpartial
The abstract and introduction clearly state the use of a two-pillar camera rig.
Verificationpartial
partial
Evidencepartial
Evaluations on 25 real-world vehicles across 10 dealerships demonstrate that our full pipeline achieves a PSNR of 28.66 dB, an SSIM of 0.89, and an LPIPS of 0.21 on held-out views
Implicationpartial
This claim is explicitly stated in the abstract with specific quantitative metrics.
Verificationpartial
partial
Evidencepartial
First, we resolve dynamic-scene ambiguities by coupling SAM 3 for instance segmentation with motion-gating to cleanly isolate the moving vehicle
Implicationpartial
The abstract clearly describes this specific step in the methodology.
Verificationpartial
partial
Evidencepartial
Second, we extract robust correspondences directly on raw, distorted 4K imagery using the RoMa v2 learned matcher guided by semantic confidence masks.
Implicationpartial
This is a detailed description of a specific method component mentioned in the abstract.
Verificationpartial
partial
Evidencepartial
Third, these matches are integrated into a rig-aware SfM optimization that utilizes CAD-derived relative pose priors to eliminate scale drift.
Implicationpartial
This describes a key aspect of the SfM process as detailed in the abstract.
Verificationpartial
partial
Evidencepartial
Finally, we use a distortion-aware 3D Gaussian Splatting framework (3DGUT) coupled with a stochastic Markov Chain Monte Carlo (MCMC) densification strategy to render reflective surfaces.
Implicationpartial
This is a precise description of the Gaussian Splatting component and its enhancement.
Verificationpartial
partial
Evidencepartial
delivering inspection-grade interactive 3D models without controlled studio infrastructure.
Implicationpartial
This is a key outcome and benefit of the proposed system, stated in the abstract.
Verificationpartial
partial
Evidencepartial
Our evaluations on 25 real-world vehicles across 10 dealerships demonstrate that our full pipeline achieves a PSNR of 28.66 dB, an SSIM of 0.89, and an LPIPS of 0.21 on held-out views
Implicationpartial
This claim is explicitly stated in the abstract with specific quantitative metrics.
Verificationpartial
partial
Evidencepartial
representing a 3.85 dB improvement over standard 3D-GS
Implicationpartial
This claim is directly stated in the abstract and supported by the comparison in Table II.
Verificationpartial
partial
Evidencepartial
First, we resolve dynamic-scene ambiguities by coupling SAM 3 for instance segmentation with motion-gating to cleanly isolate the moving vehicle, explicitly masking out non-rigid wheels to enforce strict epipolar geometry.
Implicationpartial
This claim describes a core component of the proposed method and is detailed in the abstract.
Verificationpartial
partial
Evidencepartial
Second, we extract robust correspondences directly on raw, distorted 4K imagery using the RoMa v2 learned matcher guided by semantic confidence masks.
Implicationpartial
This claim details a specific technical step in the pipeline, as described in the abstract.
Verificationpartial
partial

Author intelligence and commercialization panels stay hidden until the proof receipt is verified, cites at least 3 references, includes at least 2 sources, and clears 50% coverage. The paper narrative and citation surfaces remain public while verification is pending.

Drive-Through 3D Vehicle Exterior Reconstruction via Dynamic-Scene SfM and Distortion-Aware Gaussian Splatting

Use Signal Canvas as the narrative proof surface

Use this Signal Canvas via API or MCP

Signal Canvas proof surface