Datasets:
File size: 1,885 Bytes
5e90a14 17abce6 5e90a14 e6a3615 33e13d3 e6a3615 6f87851 e6a3615 df42789 e6a3615 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 |
---
license: cc-by-4.0
task_categories:
- object-detection
language:
- en
tags:
- Drone
- Referring Expression Comprehension
- Visual Grounding
size_categories:
- 10K<n<100K
---
# RefDrone: A Challenging Benchmark for Referring Expression Comprehension in Drone Scenes
**Images** <br>
Please download the images from [VisDrone](https://github.com/VisDrone/VisDrone-Dataset). <br>
**Annotations** <br>
Annotations is located in here. Annotation is [MDETR](https://arxiv.org/abs/2104.12763) style with keys: "images", "annotaions". <br>
#### Images
Contains information about each image in the dataset:
| Key | Description | Example |
|-----|-------------|-------------|
| `file_name` | Name of the image file | "0000189_00297_d_0000198.jpg" |
| `height` | Height of the image in pixels | 540 |
| `width` | Width of the image in pixels | 960 |
| `id` | Unique identifier for the image | 0 |
| `caption` | Text description of the image | "The black cars on the road" |
| `dataset_name` | Name of the dataset | "RefDrone" |
| `token_negative` | List of tokens not associated with target objects | [[0, 3], [4, 9]] |
#### Annotations
Contains information about the annotations for each image:
| Key | Description | Example |
|-----|-------------|-------------|
| `area` | Area of the annotated region | 1584 |
| `iscrowd` | if the annotation represents a crowd | 0|
| `image_id` | ID of the image this annotation belongs to | 0 |
| `category_id` | Category identifier for the annotated object | "5" |
| `id` | Unique identifier for the annotation | 0 |
| `empty` | if the annotation is empty | 0 |
| `bbox` | Bounding box coordinates [x, y, width, height] | [100, 200, 10, 20] |
| `tokens_positive` | List of tokens associated with the target object | [[2,4]] | |