update action task
Browse files
README.md
CHANGED
@@ -20,15 +20,29 @@ The CathAction dataset encompasses annotated frames for catheterization action u
|
|
20 |
|
21 |
These are five classes: *advance catheter*, *retract catheter*, *advance guidewire*, *retract guidewire*, and *rotate*.
|
22 |
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
|
33 |
## 2. Collision Detection
|
34 |
The CathAction dataset is designed for the collision detection task, which involves identifying whether the tip of the catheter or guidewire comes into contact with the blood vessel wall.
|
|
|
20 |
|
21 |
These are five classes: *advance catheter*, *retract catheter*, *advance guidewire*, *retract guidewire*, and *rotate*.
|
22 |
|
23 |
+
The dataset is organized into the following folders and files:
|
24 |
+
|
25 |
+
- **video_frames/**: Contains extracted video frames for each video.
|
26 |
+
- **feature_extractions/**: Contains pre-extracted RGB features, extracted using [this code](https://github.com/yjxiong/tsn-pytorch).
|
27 |
+
- **training.csv**: Groundtruth CSV file for training data.
|
28 |
+
- **validation.csv**: Groundtruth CSV file for validation data.
|
29 |
+
|
30 |
+
### Annotation File Structure
|
31 |
+
|
32 |
+
The annotation files (`training.csv` and `validation.csv`) contain four columns, with the following structure:
|
33 |
+
|
34 |
+
| Column Name | Type | Example | Description |
|
35 |
+
|---------------------|------------------|--------------|-------------------------------------------------------------------------------------------------|
|
36 |
+
| `video_id` | string | `video_1` | ID of the video where the action segment is located. |
|
37 |
+
| `start_frame` | int | `430` | Start frame of the action. |
|
38 |
+
| `stop_frame` | int | `643` | End frame of the action. |
|
39 |
+
| `all_action_classes`| list of int(s) | `[1]` | List of numeric IDs for all detected action classes in the segment. |
|
40 |
+
|
41 |
+
The frames and pre-extracted RGB features are located in the `video_frames` and `feature_extractions` folders, respectively, and can be generated using [this code](https://github.com/yjxiong/tsn-pytorch).
|
42 |
+
|
43 |
+
### Usage
|
44 |
+
|
45 |
+
1. **Catheterization Action Recognition and Anticipation Models**: Use the `start_frame` and `stop_frame` values, along with the ground truth `all_action_classes` in the CSV file, to train models that recognize action segments and anticipate future catheter actions.
|
46 |
|
47 |
## 2. Collision Detection
|
48 |
The CathAction dataset is designed for the collision detection task, which involves identifying whether the tip of the catheter or guidewire comes into contact with the blood vessel wall.
|