room_with_scan.glb: scan points as small tetrahedra (glTF POINTS render as spikes in Preview/Quick Look); README: drop mast3r_mesh

Bedroom1/Spatiallm1_MasterSlam1/room_with_scan.glb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:00e8df6aea92eaed6a226def7e43017a67ae6652d33b629893c6ce1b6dbfaac3
-size 4811848
+oid sha256:1f0eb0fddf74582e2c430175ae580c34f3e0461f2d52a1aea850864ffccdc9b3
+size 33612152

I87X/Spatiallm1_MasterSlam1/room_with_scan.glb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3d3b3f8e69d59690173f0e3555dae0dfaa5cd59a68f07a202de4d796d47000e7
-size 4847140
+oid sha256:21a8844dba95ab428d768766e81e1bafc446caf9adcbd702a7d58ac29a260979
+size 33647444

I87X/Spatiallm1_MasterSlam2/room_with_scan.glb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:112363f1f36c3b995b4dc37ae0bf67335d2995aa9497564f8aa0452804b5781f
-size 4824888
+oid sha256:0edd2020dbb330ff6983e5a2932bbb5726c4757cf4035b14d51d53ccf8b08a37
+size 33625184

P63/Spatiallm1_MasterSlam1/room_with_scan.glb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5eb44234814f4058574de30e34081e7ebcc0b3251c9fedfbbefd820f6c9b826f
-size 4814764
+oid sha256:065c2ff6e66782cda172d1560fa73b24497652395666c4c5cdb31162064e24a3
+size 33615068

P64/Spatiallm1_MasterSlam1/room_with_scan.glb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:148f20d8786773357642ac63f5cfb60eae102235b96ecdd821073342748f267f
-size 4818416
+oid sha256:e6f8ef90c0d01b85c1a0e15aa8692d4f241ad84c0251a4795712f8fccf508ddc
+size 33618716

README.md

@@ -81,7 +81,6 @@ A trailing number is a run/capture index within the room.
 | [MASt3R-SLAM](https://github.com/rmurai0610/MASt3R-SLAM) | Dense SLAM point cloud and camera trajectory from video (Colab or a GPU server; uncalibrated unless noted in Run notes) | `MasterSlam*_*` |
 | `room_plane_fit.py` (own script, in the folder) | Cleans the point cloud, aligns to gravity, fits wall planes, makes mesh, colliders and walls (Unity-ready) | `MasterSlam1_ScanVideo1/room_plane_fit_out/` |
 | `ply2obj.py` (own script) | Converts `scan.ply` to a textured mesh (`--mesh`: Poisson + baked texture) | `scan_mesh.obj/.mtl/.png` |
-| `mast3r_mesh.py` (own script, in `Bedroom1/MasterSlam1_ScanVideo1`) | Textured TSDF mesh from a MASt3R-SLAM run (keyframe depth, ICP pose refinement, TSDF fusion). `mast3r_mesh_gpu.py` is the same with TSDF fusion on CUDA | `scan_tsdf.obj/.mtl/.png` |
 | [SpatialLM](https://github.com/manycore-research/SpatialLM), run with the fork [lehoangan02/SpatialLM](https://github.com/lehoangan02/SpatialLM) (`stage3/pipeline.py`) for P64 and I87X | Room layout (walls, objects) from a point cloud | `Spatiallm1_*/room.glb`, `room_with_scan.glb` |
 | NeuCon | Mesh reconstruction (trimesh export) | `Bedroom1/NeuCon1/neucon.ply` |
 | RTAB-Map | RGB-D mapping, textured mesh | `P64/RTABMAP1/*.obj/.mtl/.jpg` |
@@ -117,7 +116,7 @@ A trailing number is a run/capture index within the room.
 | `run.log` | Run config and log |
 | `config.yaml` | MASt3R-SLAM config used (runs made on the GPU server) |
 | `calib_arcore.yaml` | Camera intrinsics passed with `--calib` (calibrated runs only) |
-| `scan_mesh.*`, `scan_tsdf.*` | Meshes from `ply2obj.py --mesh` and `mast3r_mesh.py`; `.log` is the tool output |
+| `scan_mesh.*` | Textured mesh from `ply2obj.py scan.ply scan_mesh.obj --mesh`; `scan_mesh.log` is the tool output |
 
 `MasterSlam1_ScanVideo1` also holds the scripts and outputs described above
 (`room_plane_fit.py`, `ply2obj.py`, `requirements.txt`, `room_plane_fit_out/`,
@@ -126,7 +125,8 @@ A trailing number is a run/capture index within the room.
 ### `Spatiallm1_*/`
 
 `room.glb` is the predicted layout. `room_with_scan.glb` is the layout together
-with the input scan. Folders made with the fork's `stage3/pipeline.py` also have
+with the input scan (300k scan points, each stored as a 1.5 cm coloured tetrahedron,
+because Apple Preview / Quick Look draw glTF point primitives as spikes). Folders made with the fork's `stage3/pipeline.py` also have
 `layout.txt` (SpatialLM text format, metres, z up, floor at z=0), `seeds.png`
 (top-down view of each seed), `summary.json` (settings, chosen seed, scores) and
 `pipeline.log`.