Commit ·
e339e53
1
Parent(s): e95bfeb
Update README.md
Browse files
README.md
CHANGED
|
@@ -20,3 +20,42 @@ dataset_info:
|
|
| 20 |
download_size: 3506288426
|
| 21 |
dataset_size: 11568760807
|
| 22 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
download_size: 3506288426
|
| 21 |
dataset_size: 11568760807
|
| 22 |
---
|
| 23 |
+
|
| 24 |
+
# RGB-D Salient Object Detection Dataset (RGB-D SOD)
|
| 25 |
+
|
| 26 |
+
RGB-D Salient Object Detection (RGB-D SOD) aims to detect and segment objects that *visually attract the most human interest* from a pair of color and depth images.
|
| 27 |
+
|
| 28 |
+
## Train
|
| 29 |
+
|
| 30 |
+
- COME-8K [8025 samples]
|
| 31 |
+
|
| 32 |
+
## Dev
|
| 33 |
+
|
| 34 |
+
- COME-E [4600 samples]
|
| 35 |
+
|
| 36 |
+
## Test
|
| 37 |
+
|
| 38 |
+
- Coming soon
|
| 39 |
+
|
| 40 |
+
## How to use
|
| 41 |
+
|
| 42 |
+
~~~python
|
| 43 |
+
from datasets import load_dataset
|
| 44 |
+
|
| 45 |
+
dataset = load_dataset(
|
| 46 |
+
"RGBD-SOD/rgbdsod_datasets", "v1", split="train", cache_dir="data"
|
| 47 |
+
)
|
| 48 |
+
print(dataset[0])
|
| 49 |
+
~~~
|
| 50 |
+
|
| 51 |
+
## BibTeX entry and citation info
|
| 52 |
+
|
| 53 |
+
```bibtex
|
| 54 |
+
@inproceedings{zhang2021rgb,
|
| 55 |
+
title={RGB-D saliency detection via cascaded mutual information minimization},
|
| 56 |
+
author={Zhang, Jing and Fan, Deng-Ping and Dai, Yuchao and Yu, Xin and Zhong, Yiran and Barnes, Nick and Shao, Ling},
|
| 57 |
+
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
|
| 58 |
+
pages={4338--4347},
|
| 59 |
+
year={2021}
|
| 60 |
+
}
|
| 61 |
+
```
|