honeydong49 commited on
Commit
c2cfe05
·
verified ·
1 Parent(s): 7ebb73c

Upload folder using huggingface_hub

Browse files
README.md ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: mixed-non-commercial
4
+ license_link: https://github.com/YangDongHeon/SpaceCache#license
5
+ library_name: pytorch
6
+ pipeline_tag: other
7
+ tags:
8
+ - point-cloud
9
+ - semantic-segmentation
10
+ - lidar
11
+ - streaming
12
+ - efficient-inference
13
+ - arxiv:2610.21207
14
+ ---
15
+
16
+ # SpaceCache checkpoints
17
+
18
+ Checkpoints for **SpaceCache: Intermediate Feature Reuse for Efficient Semantic
19
+ Segmentation on Point Cloud Streams**
20
+ ([paper](https://arxiv.org/abs/2610.21207),
21
+ [code](https://github.com/YangDongHeon/SpaceCache)).
22
+
23
+ SpaceCache is training-free: these are the backbone segmentation models it
24
+ runs on, Utonia and LitePT-B, trained on Waymo, nuScenes and ScanNet.
25
+
26
+ ## Usage
27
+
28
+ From the root of the [SpaceCache repository](https://github.com/YangDongHeon/SpaceCache):
29
+
30
+ ```bash
31
+ huggingface-cli download honeydong49/SpaceCache --local-dir .
32
+ ```
33
+
34
+ The files land under `exp/`, where the evaluation configs read them.
35
+
36
+ ## Checkpoints
37
+
38
+ | Backbone | Dataset | File |
39
+ |---|---|---|
40
+ | Utonia | Waymo | `exp/utonia/semseg-utonia-v1m1-5b-waymo-dec/model/model_best.pth` |
41
+ | Utonia | nuScenes | `exp/utonia/semseg-utonia-v1m1-4b-nuscenes-dec/model/model_best.pth` |
42
+ | Utonia | ScanNet | `exp/utonia/semseg-utonia-v1m1-0b-scannet-dec/model/model_best.pth` |
43
+ | LitePT-B | Waymo | `exp/litept/waymo/semseg-litept-v1m1-0-base/model/model_best.pth` |
44
+ | LitePT-B | nuScenes | `exp/litept/nuscenes/semseg-litept-v1m1-0-base/model/model_best.pth` |
45
+ | LitePT-B | ScanNet | `exp/litept/scannet/semseg-litept-v1m1-0-base/model/model_best.pth` |
46
+
47
+ Each file holds the model weights (`state_dict`) only. The Utonia models are
48
+ a segmentation decoder trained on top of the released
49
+ [Utonia](https://huggingface.co/Pointcept/Utonia) encoder, which was kept
50
+ frozen; the file contains both. The LitePT-B models were trained from scratch
51
+ with the [LitePT](https://github.com/prs-eth/LitePT) architecture.
52
+
53
+ ## License
54
+
55
+ All checkpoints are for non-commercial research only.
56
+
57
+ | Checkpoints | Terms |
58
+ |---|---|
59
+ | Utonia (all datasets) | contain the [Utonia](https://huggingface.co/Pointcept/Utonia) weights, CC BY-NC 4.0, plus the terms of the dataset below |
60
+ | Waymo | [Waymo Dataset License Agreement for Non-Commercial Use](https://waymo.com/open/terms) (notice below) |
61
+ | nuScenes | CC BY-NC-SA 4.0, [nuScenes terms of use](https://www.nuscenes.org/terms-of-use) |
62
+ | ScanNet | [ScanNet Terms of Use](http://kaldir.vc.in.tum.de/scannet/ScanNet_TOS.pdf) |
63
+
64
+ The checkpoints trained on Waymo were made using the Waymo Open Dataset,
65
+ provided by Waymo LLC under the Waymo Dataset License Agreement for
66
+ Non-Commercial Use, available at waymo.com/open/terms, and your access and use
67
+ of such work are governed by the terms and conditions therein. They are
68
+ released under that agreement; any further downstream use or modification of
69
+ them is subject to its terms, including the non-commercial restrictions of its
70
+ Section 4. They must not be used to operate or assist the operation of a
71
+ vehicle, in production systems, or for other primarily commercial purposes.
72
+
73
+ ## Citation
74
+
75
+ ```bibtex
76
+ @article{yang2026spacecache,
77
+ title = {SpaceCache: Intermediate Feature Reuse for Efficient Semantic Segmentation on Point Cloud Streams},
78
+ author = {Yang, Dongheon and Kang, Beomseok and Kim, Jae-Joon},
79
+ journal = {arXiv preprint arXiv:2610.21207},
80
+ year = {2026}
81
+ }
82
+ ```
exp/litept/nuscenes/semseg-litept-v1m1-0-base/model/model_best.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee622e1255163696f342a92227cad0dfc172222e771c385b240d7013a97020dc
3
+ size 180565602
exp/litept/scannet/semseg-litept-v1m1-0-base/model/model_best.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:399678d9034b992bc94107a6089e0230b7b1cbf1b5d52e43bebbddb1eb8e0feb
3
+ size 180620770
exp/litept/waymo/semseg-litept-v1m1-0-base/model/model_best.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:413435edec10ddc52ebfb2beb2f86b68718fa2eddd04642ae3e037706eece143
3
+ size 180567330
exp/utonia/semseg-utonia-v1m1-0b-scannet-dec/model/model_best.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:001ba8cc5645eb4afda898f3fb81519c2905b41745bc7c913bbc1b15487ee243
3
+ size 631206484
exp/utonia/semseg-utonia-v1m1-4b-nuscenes-dec/model/model_best.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1770020409e7f8bc0d30839e3023e80d06cf16ee3aaad52919a4fcc0eb0e7b29
3
+ size 631205652
exp/utonia/semseg-utonia-v1m1-5b-waymo-dec/model/model_best.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c0962fdf072ffbf57fa8fb38280a10150a7689d0eac16490952be8aca743157
3
+ size 631206932