Datasets:

Modalities:
Image
ArXiv:
License:
JeffreyJsam commited on
Commit
831376d
Β·
verified Β·
1 Parent(s): 58702f3

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +64 -3
README.md CHANGED
@@ -1,3 +1,64 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - image
5
+ - segmentation
6
+ - space
7
+ pretty_name: 'SWiM: Spacecraft With Masks (Instance Segmentation)'
8
+ size_categories:
9
+ - 1K<n<1M
10
+ task_categories:
11
+ - image-segmentation
12
+ task_ids:
13
+ - instance-segmentation
14
+ annotations_creators:
15
+ - machine-generated
16
+ - expert-generated
17
+ ---
18
+
19
+ ---
20
+
21
+ # SWiM: Spacecraft With Masks
22
+
23
+ A large-scale instance segmentation dataset of nearly 64k annotated spacecraft images that was created using real spacecraft models, superimposed on a mixture of real and synthetic backgrounds generated using NASA's TTALOS pipeline. To mimic camera distortions and noise in real-world image acquisition, we also added different types of noise and distortion to the images.
24
+
25
+ ## Dataset Summary
26
+ The dataset contains over 64,000 annotated images with instance masks for varied spacecraft. It's structured for YOLO and segmentation applications, and chunked to stay within Hugging Face's per-folder file limits.
27
+
28
+
29
+ ## How to Use
30
+ ### Directory Structure Note
31
+
32
+ Due to Hugging Face Hub's per-directory file limit (10,000 files), this dataset is chunked: each logical split (like `train/labels/`) is subdivided into folders (`000/`, `001/`, ...) containing no more than 5,000 files each.
33
+
34
+ **Example Structure:**
35
+
36
+ labels/
37
+ β”œβ”€β”€ 000/
38
+ β”‚ β”œβ”€β”€ img_0.png
39
+ β”‚ └── ...
40
+ β”œβ”€β”€ 001/
41
+ └── ...
42
+ If you're using models/tools like **YOLO** or others that expect a **flat directory**, you may need to **merge these subfolders at load-time or during preprocessing**.
43
+
44
+
45
+ ## Code and Data Generation Pipeline
46
+
47
+ All dataset generation scripts, preprocessing tools, and model training code are available on GitHub:
48
+
49
+ [GitHub Repository: https://github.com/RiceD2KLab/SWiM](https://github.com/RiceD2KLab/SWiM)
50
+
51
+
52
+ ## Citation
53
+
54
+ If you use this dataset, please cite:
55
+
56
+ @misc{sam2025newdatasetperformancebenchmark,
57
+ title={A New Dataset and Performance Benchmark for Real-time Spacecraft Segmentation in Onboard Flight Computers},
58
+ author={Jeffrey Joan Sam and Janhavi Sathe and Nikhil Chigali and Naman Gupta and Radhey Ruparel and Yicheng Jiang and Janmajay Singh and James W. Berck and Arko Barman},
59
+ year={2025},
60
+ eprint={2507.10775},
61
+ archivePrefix={arXiv},
62
+ primaryClass={cs.CV},
63
+ url={https://arxiv.org/abs/2507.10775},
64
+ }