Research map / 3D vision / NeRF and radiance fields
LaLaLoc++: Global Floor Plan Comprehension for Layout Localisation in Unvisited Environments
Henry Howard-Jenkins, Victor Adrian Prisacariu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace LaLaLoc with LaLaLoc++ (dense floor plan comprehension + transformer) | floor plan localisation · Structured3D | median error (cm): 8.3 -> 5.2 | Improved |
| replace LaLaLoc with LaLaLoc++ | floor plan localisation · Structured3D | accuracy <1m: 87.5% -> 92.3% | Improved |
| replace LaLaLoc with LaLaLoc++ | floor plan localisation · Zillow Indoor Dataset | median error (cm): 13.1 -> 9.3 | Improved |
| replace rendering-based LaLaLoc with direct 2D plan inference | LaLaLoc · Structured3D inference time | total inference time (s): 4.65 -> 0.13 | Improved |
| train Phi_plan to mimic LaLaLoc layout branch (ablation) | LaLaLoc++ Phi_plan · Structured3D | median error (cm): 8.3 -> 8.4 | No clear effect |
| train Phi_plan with layout decoding loss instead of mimicking (ablation) | LaLaLoc++ Phi_plan · Structured3D | median error (cm): 8.4 -> 5.7 | Improved |
| add transformer encoder to the image embedding branch (ablation) | LaLaLoc++ Phi_image · Structured3D | median error (cm): 5.7 -> 5.2 | Improved |
| use full-size latent floor plan instead of 8x subsampling (ablation) | LaLaLoc++ refinement · Structured3D | median error (cm): 6.5 -> 5.2 | Improved |
| use full-size latent floor plan instead of 4x subsampling (ablation) | LaLaLoc++ refinement · Structured3D | median error (cm): 5.3 -> 5.2 | No clear effect |
| use 5 down/up-sample layers in Phi_plan instead of 3 (ablation) | LaLaLoc++ Phi_plan · Structured3D | median error (cm): 12.9 -> 5.2 | Improved |
| use 5 layers instead of 4 in Phi_plan (ablation) | LaLaLoc++ Phi_plan · Structured3D | median error (cm): 6.0 -> 5.2 | Improved |
| use 6 layers instead of 5 in Phi_plan (ablation) | LaLaLoc++ Phi_plan · Structured3D | median error (cm): 5.2 -> 5.3 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- LaLaLoc — LaLaLoc: Latent Layout Localisation in Dynamic, Unvisited Environments (ICCV 2021) · better on 11 of 11 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- LaLaLoc: Latent Layout Localisation in Dynamic, Unvisited Environments (ICCV 2021 · similarity 0.93)
- HouseLayout3D: A Benchmark and Training-free Baseline for 3D Layout Estimation in the Wild (NeurIPS 2025 · similarity 0.88)
- LayoutVLM: Differentiable Optimization of 3D Layout via Vision-Language Models (CVPR 2025 · similarity 0.86)
- OptiScene: LLM-driven Indoor Scene Layout Generation via Scaled Human-aligned Data Synthesis and Multi-Stage Preference Optimization (NeurIPS 2025 · similarity 0.86)
- MvFFN: Multi-view Floor-Plan Feed-Forward Network for Unposed Wide-Baseline Panorama Layout Reconstruction (NeurIPS 2026 · similarity 0.86)
- F3Loc: Fusion and Filtering for Floorplan Localization (CVPR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
