​

Tabular Formats: Parquet / CSV

​

Image Annotations

For full-image class labels, create an annotation file named image_annotations.csv orimage_annotations.parquet. Each row in the file represents an image and its corresponding label.

​

Example Format:

• The filename column must contain relative paths.
• The label column assigns a class to entire image.
• Multiple labels can be stored as a list, e.g., ['t-shirt', 'SKU12345'].
• You may also include a caption column for textual metadata.

​

Example for Multiple Labels:

​

Object Annotations

For object-level annotations, create an annotation file named object_annotations.csv or object_annotations.parquet. Each row represents a detected object, including its bounding box coordinates and class label.

​

Example Format:

• col_x and row_y define the top-left corner of the bounding box.
• width and height must be greater than zero.
• Each row corresponds to a single object within an image.

​

JSON Annotations (COCO Format)

Ensure The file is named annotations.json.

​

Example JSON Format:

{
    "images": [
        { "id": 1, "width": 640, "height": 480, "file_name": "image1.jpg" },
        { "id": 2, "width": 800, "height": 600, "file_name": "image2.jpg" }
    ],
    "categories": [
        { "id": 1, "name": "cat" },
        { "id": 2, "name": "dog" },
        { "id": 3, "name": "t-rex" }
    ],
    "annotations": [
        { "id": 1, "image_id": 1, "category_id": 1, "bbox": [100, 100, 200, 200] },
        { "id": 2, "image_id": 2, "category_id": 2, "bbox": [50, 50, 150, 150] },
        { "id": 3, "image_id": 1, "category_id": 3 },
        { "id": 4, "image_id": 2, "category_id": 3 }
    ]
}

• Bounding boxes (bbox) follow the format [col_x, row_y, width, height].
  • col_x and row_y define the top-left corner of the bounding box.
  • width and height must be greater than zero.
• Remove comments before uploading.

​

Summary of Bounding Box Formats

For instructions on importing annotation files, see Importing Annotations.

By following these guidelines, you can ensure that your annotation files are structured correctly and ready for seamless dataset creation in Visual Layer.