Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
image
imagewidth (px)
1.92k
1.92k
label
class label
3 classes
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
0Test
End of preview. Expand in Data Studio

3DCuboidOnRoad

3DCuboidOnRoad is a road-scene dataset with vehicle annotations in the CVAT 1.1 cuboid format. The dataset contains images of cars, buses, and trucks annotated with 3D-style cuboids.

The README is available in two languages:


Russian Version

Описание

3DCuboidOnRoad - это набор данных с изображениями дорожных сцен и аннотациями транспортных средств в виде кубоидов (cuboid).

Разметка выполнена в CVAT в формате CVAT 1.1. Кубоиды были созданы с помощью проекта Auto-3D-cuboids-for-CVAT.

Набор данных предназначен для задач компьютерного зрения, связанных с дорожными сценами и транспортными средствами.

Возможные направления использования:

  • обнаружение транспортных средств;
  • 3D-style object detection;
  • локализация объектов по кубоидам;
  • анализ дорожных сцен;
  • исследования в области автономного вождения;
  • обучение и тестирование моделей компьютерного зрения.

Структура набора данных

Набор данных содержит 2000 изображений и разделён на три выборки:

Выборка Количество изображений
Train 1400
Valid 300
Test 300

Структура файлов:

3DCuboidOnRoad/
+-- image/
|   +-- Train/
|   +-- Valid/
|   +-- Test/
+-- annotations.xml
+-- README.md

Формат аннотаций

Аннотации находятся в файле:

annotations.xml

Формат разметки:

CVAT for images 1.1

Каждый объект размечен как cuboid. В XML используются координаты кубоида, которые поддерживаются CVAT для 2D-проекции 3D-объекта.

Классы

В наборе данных три класса:

Имя класса Количество объектов Процент от общего
Car 11499 92.99%
Bus 63 0.51%
Truck 804 6.50%

Всего объектов:

12366

Как использовать в CVAT

  1. Создайте задачу в CVAT и загрузите изображения из папки image/.
  2. Убедитесь, что имена изображений совпадают с именами в annotations.xml.
  3. Импортируйте annotations.xml.
  4. Выберите формат импорта CVAT for images 1.1.

Автор

Автор набора данных: Зейналов У.Р.о.


English Description

Description

3DCuboidOnRoad is a dataset of road-scene images with vehicle annotations represented as cuboids (cuboid).

The annotations were created in CVAT using the CVAT 1.1 format. The cuboids were generated with the Auto-3D-cuboids-for-CVAT project.

The dataset is intended for computer vision tasks related to road scenes and vehicles.

Possible use cases:

  • vehicle detection;
  • 3D-style object detection;
  • object localization using cuboids;
  • road-scene analysis;
  • autonomous driving research;
  • training and evaluation of computer vision models.

Dataset Structure

The dataset contains 2000 images and is split into three subsets:

Split Number of images
Train 1400
Valid 300
Test 300

File structure:

3DCuboidOnRoad/
+-- image/
|   +-- Train/
|   +-- Valid/
|   +-- Test/
+-- annotations.xml
+-- README.md

Annotation Format

The annotation file is:

annotations.xml

Annotation format:

CVAT for images 1.1

Each object is annotated as a cuboid. The XML file contains cuboid coordinates supported by CVAT for 2D projections of 3D-style objects.

Classes

The dataset contains three object classes:

Class name Number of objects Percentage of total
Car 11499 92.99%
Bus 63 0.51%
Truck 804 6.50%

Total number of objects:

12366

How to Use in CVAT

  1. Create a CVAT task and upload the images from the image/ directory.
  2. Make sure that image names match the names stored in annotations.xml.
  3. Import annotations.xml.
  4. Select the CVAT for images 1.1 import format.

Author

Dataset author: Zeynalov U.R.O.

Downloads last month
1,430