Paper · ≈ 1 min read
latentSplat: Autoencoding Variational Gaussians for Fast Generalizable 3D Reconstruction
Outgoing relations
- belongs_to_approach_family → Gaussian Splatting Reconstruction (ApproachFamily)
- part_of → 3D/4D Vision and Reconstruction (ResearchField)
- uses → Gaussian Splatting (MethodologicalConcept)
- uses → Multi-View Consistency (EvaluationProtocol)
- uses_metric → LPIPS (Metric)
- uses_metric → PSNR (Metric)
- uses_metric → SSIM (Metric)
Connected node types
- Metric: 3
- ApproachFamily: 1
- EvaluationProtocol: 1
- MethodologicalConcept: 1
- ResearchField: 1
Mentions in the corpus
- part_of Evidence: --- type: Paper arxiv: "2403.16292" arxiv_url: https://arxiv.org/abs/24…
- mentioned_in Claim: --- type: Paper arxiv: "2403.16292" arxiv_url: https://arxiv.org/abs/2403.16292 title: "latentS…
- part_of Evidence: We present latentSplat, a method to predict semantic Gaussians in a 3D…
- part_of Evidence: We show that latentSplat outperforms previous works in reconstruction q…
- part_of Evidence: Training relies entirely on readily available real video data without e…
- mentioned_in Claim: Training relies entirely on readily available real video data without explicit 3D supervision,…
- part_of Evidence: Experiments show that latentSplat outperforms previous generalisable me…
- summarizes Project Pulse
- synthesizes Topic — Gaussian Splatting Reconstruction
- summarizes Topic — Gaussian Splatting Reconstruction
- synthesizes Comparison — Dynamic Gaussian Splatting vs Gaussian Splatting Reconstruction
- summarizes Comparison — Dynamic Gaussian Splatting vs Gaussian Splatting Reconstruction
- summarizes Field Overview — 3D/4D Vision and Reconstruction
Cross-references in raw data
Source provenance
examples/demo-corpus/data/research/papers/arxiv-2403-16292/abstract.md