same-hardware latency framing (138x/65x re-timed) + replace stale p>0.75 with effect-size framing

README.md

@@ -14,7 +14,7 @@ library_name: pytorch
 
 # ScaffDiff: Scaffold-Dominant Diffusion for 3D Scene Completion
 
-Single-step, scaffold-dominant diffusion for 3D scene completion on SemanticKITTI. Released alongside two paper submissions: an SMC 2026 short paper on the multi-token Gaussian VAE and ICP-refined ground truth (v2 GT), and an RA-L 2026 journal paper on scaffold-dominance and a 45-minute mixed-scaffold deployment fine-tune. The pipeline runs at 209 ms/frame (4.78 FPS) on a single RTX 4090 — 143× faster than LiDiff and 25× faster than ScoreLiDAR — and the fine-tuned teacher beats both by 69–72 % squared Chamfer in a paired matched-protocol scaffold-free comparison across 6 fine-tune seeds.
+Single-step, scaffold-dominant diffusion for 3D scene completion on SemanticKITTI. Released alongside two paper submissions: an SMC 2026 short paper on the multi-token Gaussian VAE and ICP-refined ground truth (v2 GT), and an RA-L 2026 journal paper on scaffold-dominance and a 45-minute mixed-scaffold deployment fine-tune. The pipeline runs at 209 ms/frame (4.78 FPS) on a single RTX 4090 (138× faster than LiDiff and 65× faster than ScoreLiDAR, re-timed on the same hardware; 143×/25× against author-reported timings) and the fine-tuned teacher beats both by 69–72 % squared Chamfer in a paired matched-protocol scaffold-free comparison across 6 fine-tune seeds.
 
 > Note on the previous version of this card. The earlier README framed this work around "modality-agnostic encoder" and "teacher-vs-student distillation," with student CDs reported on a different metric protocol. Those numbers and that framing are superseded by the two paper drafts that this card now mirrors. See [What changed vs. the old card](#what-changed-vs-the-old-card) at the bottom.
 
@@ -33,7 +33,7 @@ The journal paper extends the analysis with:
 - **Scaffold-dominance ablations** (six controlled runs): zeroing the 108 M PTv3 encoder costs only +58 % CD²; cutting input density from 20 K to 2 K points costs only +2 %; but replacing the GT scaffold with random noise or a regular voxel grid collapses CD² by ≥4 orders of magnitude (>50,000× and ~43,000× respectively).
 - **Random-init PTv3 control**: removing the 108 M parameters of self-supervised pretraining does not open a LiDAR↔DA2 gap (CD² 0.0279 vs 0.0279 m²). The frozen encoder is therefore not the load-bearing component — the GT coordinate scaffold is.
 - **45-minute mixed-scaffold fine-tune**: 3-epoch fine-tune (lr 5e-5, 30 % LIDAR-cropped scaffold mix, t ∈ [50, 400], encoder frozen) that transfers the same denoiser to a deployment-realistic scaffold (per-frame LiDAR sweep, ego-bbox crop, **no GT access**) at squared CD **0.968 ± 0.194 m²** across 6 fine-tune seeds on a paired 50-frame protocol — **69–72 % below LiDiff / ScoreLiDAR**, 13× over the pre-FT baseline.
-- **Cross-sequence in-distribution evaluation** on seqs 00 / 05 / 08 (range 6.7 % in CD²), DDIM multi-step sweep (single-step at t=200 is optimal), 500-frame scaffold-quality jitter sweep, iterative self-scaffolding drift study, and an N-seed ensemble that confirms the LiDAR↔DA2 gap stays within sampling noise (max |ΔCD²| ≤ 7×10⁻⁴, paired Wilcoxon p > 0.75) at every N ∈ {1, 2, 4, 8}.
+- **Cross-sequence in-distribution evaluation** on seqs 00 / 05 / 08 (range 6.7 % in CD²), DDIM multi-step sweep (single-step at t=200 is optimal), 500-frame scaffold-quality jitter sweep, iterative self-scaffolding drift study, and an N-seed ensemble that confirms the LiDAR↔DA2 gap stays within sampling noise (max |ΔCD²| ≤ 7×10⁻⁴, within sampling noise) at every N ∈ {1, 2, 4, 8}.
 
 ---
 
@@ -59,7 +59,7 @@ All numbers are squared symmetric Chamfer distance in m², matching the PVD / Li
 | Ours (teacher, GT-scaffold OOD here)    | single-step x₀                   | 12.58 ± 8.14      |
 | **Ours (teacher-FT, mixed scaffold, 6 seeds)** | **single-step x₀ (kdtree match)** | **0.968 ± 0.194** |
 
-**69–72 % below LiDiff / ScoreLiDAR** (6-seed across-seed mean ± std, epoch-2 selection; per-seed kdtree CD² 0.73 / 1.03 / 0.75 / 1.12 / 0.97 / 1.21). Paired Wilcoxon p < 1e-12 (seed 42 reference). End-to-end latency is unchanged at 209 ms/frame: 143× faster than LiDiff (30 s/frame, 50 steps), 25× faster than ScoreLiDAR (5.37 s/frame, 8 steps). Non-diffusion LiNeXt (167 ms) is the only comparable-latency baseline.
+**69–72 % below LiDiff / ScoreLiDAR** (6-seed across-seed mean ± std, epoch-2 selection; per-seed kdtree CD² 0.73 / 1.03 / 0.75 / 1.12 / 0.97 / 1.21). Paired Wilcoxon p < 1e-12 (seed 42 reference). End-to-end latency is unchanged at 209 ms/frame: 138× faster than LiDiff (28.95 s/frame re-timed on our RTX 4090; 30 s as reported), 65× faster than ScoreLiDAR (13.61 s/frame re-timed; 5.37 s as reported). Non-diffusion LiNeXt (167 ms) is the only comparable-latency baseline.
 
 ### C) Cross-sequence in-distribution evaluation
 | Sequence | n frames | CD² (m²)         | F@0.2 | H₉₅ (m) |
@@ -77,7 +77,7 @@ LiDAR vs DA2 (Depth Anything V2 monocular pseudo-LiDAR):
 - JSD: 0.0337 vs 0.0337
 - F@0.2: 0.8439 vs 0.8435
 - H₉₅: 0.2854 vs 0.2863 m
-- |ΔCD²| < 4×10⁻⁴ m²; Wilcoxon p > 0.75. Result also holds with a **randomly initialised** PTv3 encoder.
+- |ΔCD²| < 4×10⁻⁴ m², an order of magnitude below the encoder-zero gap. Result also holds with a **randomly initialised** PTv3 encoder.
 
 ### E) VAE reconstruction (SMC paper)
 | Metric | Value |