insailab commited on
Commit
9c5308d
·
verified ·
1 Parent(s): 6bfc5a4

docs: add DELIVER dataset card

Browse files
Files changed (1) hide show
  1. README.md +119 -0
README.md ADDED
@@ -0,0 +1,119 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ annotations_creators:
3
+ - expert-generated
4
+ language_creators:
5
+ - crowdsourced
6
+ language:
7
+ - en
8
+ license: apache-2.0
9
+ multilinguality:
10
+ - monolingual
11
+ size_categories:
12
+ - 10K<n<100K
13
+ source_datasets:
14
+ - original
15
+ task_categories:
16
+ - image-segmentation
17
+ task_ids:
18
+ - semantic-segmentation
19
+ pretty_name: DeLiVER
20
+ tags:
21
+ - multimodal
22
+ - autonomous-driving
23
+ - lidar
24
+ - event-camera
25
+ - depth
26
+ - rgb
27
+ - semantic-segmentation
28
+ - cvpr-2023
29
+ ---
30
+
31
+ # DeLiVER: Delivering Arbitrary-Modal Semantic Segmentation
32
+
33
+ [CVPR 2023] Official Dataset Repository for the **DeLiVER** benchmark.
34
+
35
+ - **Project Page:** https://jamycheung.github.io/DELIVER.html
36
+ - **GitHub Repository:** https://github.com/InSAI-Lab/DELIVER
37
+ - **Paper:** [arXiv:2303.01480](https://arxiv.org/abs/2303.01480)
38
+ - **Model Checkpoints:** [InSAI-Lab/CMNeXt](https://huggingface.co/InSAI-Lab/CMNeXt)
39
+
40
+ ---
41
+
42
+ ## Dataset Description
43
+
44
+ To conduct arbitrary-modal semantic segmentation, we created the **DeLiVER** benchmark, covering **De**pth, **Li**DAR, multiple **V**iews, **E**vents, and **R**GB.
45
+ It includes 4 adverse weather conditions (**cloudy, foggy, night-time, sunny**) and 5 sensor failure corner cases:
46
+ - **MB:** Motion Blur
47
+ - **OE:** Over-Exposure
48
+ - **UE:** Under-Exposure
49
+ - **LJ:** LiDAR-Jitter
50
+ - **EL:** Event Low-resolution
51
+
52
+ Each sample has 6 views, 4 modalities, and 2 annotation types (semantic and instance).
53
+
54
+ ## File Structure
55
+
56
+ The dataset archive `DELIVER.tar.gz` (~12.2 GB) unpacks into:
57
+
58
+ ```text
59
+ DELIVER/
60
+ ├── depth/
61
+ │ ├── cloud/ (test, train, val)
62
+ │ ├── fog/
63
+ │ ├── night/
64
+ │ ├── rain/
65
+ │ └── sun/
66
+ ├── event/
67
+ ├── hha/
68
+ ├── img/
69
+ ├── lidar/
70
+ └── semantic/
71
+ ```
72
+
73
+ ## Quick Start & Download
74
+
75
+ Using the Hugging Face CLI:
76
+
77
+ ```bash
78
+ # Download dataset archive
79
+ hf download InSAI-Lab/DELIVER DELIVER.tar.gz --type dataset --local-dir ./data
80
+
81
+ # Extract archive
82
+ cd ./data
83
+ tar -zxvf DELIVER.tar.gz
84
+ ```
85
+
86
+ Using Python (`huggingface_hub`):
87
+
88
+ ```python
89
+ from huggingface_hub import hf_hub_download
90
+
91
+ dataset_path = hf_hub_download(
92
+ repo_id="InSAI-Lab/DELIVER",
93
+ repo_type="dataset",
94
+ filename="DELIVER.tar.gz",
95
+ local_dir="./data"
96
+ )
97
+ ```
98
+
99
+ ## Citation
100
+
101
+ ```bibtex
102
+ @inproceedings{zhang2023delivering,
103
+ title={Delivering Arbitrary-Modal Semantic Segmentation},
104
+ author={Zhang, Jiaming and Liu, Ruiping and Shi, Hao and Yang, Kailun and Rei{\ss}, Simon and Peng, Kunyu and Fu, Haodong and Wang, Kaiwei and Stiefelhagen, Rainer},
105
+ booktitle={CVPR},
106
+ year={2023}
107
+ }
108
+
109
+ @article{zhang2023cmx,
110
+ title={CMX: Cross-modal fusion for RGB-X semantic segmentation with transformers},
111
+ author={Zhang, Jiaming and Liu, Huayao and Yang, Kailun and Hu, Xinxin and Liu, Ruiping and Stiefelhagen, Rainer},
112
+ journal={IEEE Transactions on Intelligent Transportation Systems},
113
+ year={2023}
114
+ }
115
+ ```
116
+
117
+ ## License
118
+
119
+ This dataset is released under the [Apache-2.0 License](https://www.apache.org/licenses/LICENSE-2.0).