Add dataset card and task categories

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +67 -3
README.md CHANGED
@@ -1,3 +1,67 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ task_categories:
4
+ - image-segmentation
5
+ tags:
6
+ - remote-sensing
7
+ - open-vocabulary
8
+ - earth-observation
9
+ ---
10
+
11
+ # OVRSIS95K
12
+
13
+ This dataset is part of the **OVRSISBenchV2** benchmark presented in the paper [Towards Realistic Open-Vocabulary Remote Sensing Segmentation: Benchmark and Baseline](https://huggingface.co/papers/2604.15652).
14
+
15
+ Official Code Repository: [LiBingyu01/Pi-Seg](https://github.com/LiBingyu01/Pi-Seg)
16
+
17
+ ## Dataset Overview
18
+
19
+ **OVRSIS95K** is a large-scale, balanced training dataset designed for open-vocabulary remote sensing image segmentation (OVRSIS). It contains approximately 95,000 image-mask pairs covering 35 common semantic categories across diverse remote sensing scenes.
20
+
21
+ ### Key Highlights
22
+ - **~95K high-quality annotated image-mask pairs**
23
+ - **35 semantic categories** with a balanced class distribution to mitigate long-tail challenges.
24
+ - **5 scene domains**: town, industrial, forest, waterfront, and wasteland.
25
+ - Serves as the training foundation for the **OVRSISBenchV2** benchmark.
26
+
27
+ ## Dataset Structure
28
+
29
+ The recommended dataset directory structure is:
30
+
31
+ ```text
32
+ datasets/
33
+ └── OVRSIS95K/
34
+ ├── train/
35
+ │ ├── images/
36
+ │ └── annotations/
37
+ └── val/
38
+ ├── images/
39
+ └── annotations/
40
+ ```
41
+
42
+ ## Related Datasets
43
+ - **OVRSISBenchV2 (OVRSIS test set)**: [kkk2026/OVRSISBenchtest](https://huggingface.co/datasets/kkk2026/OVRSISBenchtest)
44
+ - **OVRSISBenchV2 (3 downstream tasks)**: [kkk2026/OVRSISBenchV2o3](https://huggingface.co/datasets/kkk2026/OVRSISBenchV2o3)
45
+
46
+ ## Citation
47
+
48
+ If you use this dataset or the associated code, please cite the following papers:
49
+
50
+ ```bibtex
51
+ @article{li2026towards,
52
+ title={Towards Realistic Open-Vocabulary Remote Sensing Segmentation: Benchmark and Baseline},
53
+ author={Li, Bingyu and Huo, Tao and Dong, Haocheng and Zhang, Da and Zhao, Zhiyuan and Gao, Junyu and Li, Xuelong},
54
+ journal={arXiv preprint arXiv:2604.15652},
55
+ year={2026}
56
+ }
57
+
58
+ @inproceedings{li2026exploring,
59
+ title={Exploring efficient open-vocabulary segmentation in the remote sensing},
60
+ author={Li, Bingyu and Dong, Haocheng and Zhang, Da and Zhao, Zhiyuan and Sun, Hao and Gao, Junyu},
61
+ booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
62
+ volume={40},
63
+ number={8},
64
+ pages={5982--5991},
65
+ year={2026}
66
+ }
67
+ ```