xuyankun commited on
Commit
5b15c8a
1 Parent(s): 3dc7934

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +9 -2
README.md CHANGED
@@ -6,7 +6,14 @@ size_categories:
6
 
7
  This is the official website for raw video dataset used in the paper [VSViG: Real-time Video-based Seizure Detection via Skeleton-based Spatiotemporal ViG](https://arxiv.org/pdf/2311.14775.pdf), which is accepted by ECCV2024.
8
 
9
- The detailed data description and annotation will be added soon!
10
 
11
- The project homepage is available at [here](https://github.com/xuyankun/VSViG).
12
 
 
 
 
 
 
 
 
 
6
 
7
  This is the official website for raw video dataset used in the paper [VSViG: Real-time Video-based Seizure Detection via Skeleton-based Spatiotemporal ViG](https://arxiv.org/pdf/2311.14775.pdf), which is accepted by ECCV2024.
8
 
9
+ The `Label.xlsx` file shows labels about time of EEG onset and clnical onset annotated by clinical experts for each seizure.
10
 
11
+ EEG onset means abnormalities appear in EEG recodrings, sometimes it cannot cause behavioral symptoms, while clinical onset is defined by obvious behavioral abnoramlities can be detected.
12
 
13
+ The seizure type _P_ means partial seizure, known as focal seizure, and _PG_ means Partial to Generalized seizure which is more severe and lasts longer.
14
+
15
+ In order to preserve pravicy of patients and their families, we have to mask patients' face. However, this operation leads to sutle characteristcs/motions on face, which is hardly to detect the seizure onset early. We present the EEG onset and clinical onset time based on our raw non-mosaic video, so if you want to use this dataset, you can redefine the EEG/clinical onset based on your scenarios.
16
+
17
+ Last, we will add more patients in the future.
18
+
19
+ The project homepage is available at [here](https://github.com/xuyankun/VSViG).