# A02025-Object-Detection — Data

来源：https://www.kaggle.com/competitions/a0-2025-object-detection/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

The dataset is organized into two main folders: **Train** and **Test**. The structure of each folder is as follows:

- **Train** folder: divided into two subfolders:
  - **images**: Contains images file.
  - **labels**: Contains labels file, each including the classification label (Apple: 1, Grapes: 2, Pineapple: 3, Orange: 4, Banana: 5, and Watermelon: 6) and the bounding box coordinates.
  
- **Test** folder: Contains images without labels. These images are used for both the **public test** and the **private test**.
  - **Public Test**: Results from the public test will be displayed on the public leaderboard after submission.
  - **Private Test**: Participants must select one of their previous submissions for evaluation on the private test. The final ranking will be determined based on the private test, and results will be displayed on the private leaderboard after the contest concludes.

Each image in the dataset is in .jpg format. The **Train** and **Val** folders contain labeled data, allowing participants to train and validate their models, while the **Test** folder is used for final evaluation without access to labels.