Smart Uploader (beta)

Overview

The Smart Uploader allows users to upload large sets of images while automatically parsing metadata from filenames. By applying custom parsing patterns, the system groups images by subjects and associates metadata fields such as subject ID, side (left/front/right), and additional study attributes.

This feature is designed to save time, reduce manual work, and ensure consistent image organization for datasets.

Key Capabilities

• Dataset selection: Upload images into an existing dataset.

• Filename parsing: Extract metadata from filenames using standard or custom parsing rules.

• Automatic grouping: Images are grouped by subject and side for analysis.

• Custom metadata fields: Define additional fields (e.g., timeCode, productCode, light, studyNumber, etc.) that are stored automatically in metadata.

• Flexible upload: Supports drag & drop or file browsing for PNG, JPEG, and WebP images (max 30MB per file).

Filename Notation Rules

Standard Fields

• subject (required): Unique subject identifier.

• side (optional): Defines the camera angle or view. Possible values: Left, Front (default if not set), or Right. If the side parameter is used, the order must follow: Left → Front → Right.

Custom Fields

Any other field can be included as a metadata field:

• timeCode: Defines the order of photos. Use this field to specify the sequence in which photos should be processed or displayed.

• productCode

• light

• studyNumber

• areaCode

• …and more. These will be stored as custom metadata.

Pattern Syntax

Use angle brackets < > to define a parsing rule.

Examples:

• Field length: <4=subject> → First 4 characters as subject ID.

• Field name: <subject>, <side>, <timeCode>

• Value mapping (for side images):

  • ('L'|'F'|'R') → Left, Front, Right

  • ('LeftView'|'FrontView'|'RightView')

Basic Format

<length=fieldName>_<side('L'|'F'|'R')>

• Combine multiple segments with separators: _, -, space.

• camelCase and PascalCase are also supported.

Examples

• 001_F.jpg : <3=subject>_<side('L'|'F'|'R')>

• P123-Front-001.png : <productCode>-<side('Left'|'Front'|'Right')>-<3=subject>

Notes & Best Practices

• Always include subject and side in your pattern.

• Use fixed lengths for subject IDs when possible (e.g., <3=subject>).

• Use value lists for fields with known options (side, light).

• Test your patterns on a few filenames before batch uploads.

• Use consistent separators (_, -, space) across filenames.

Supported Formats

• Image types: PNG, JPEG, WebP

• Max size: 30MB per image

Photo Statuses

• Match: The photo name matches the required pattern/regular expression and is parsed successfully.

• Ready: The photo has (identified via pattern matching or manually) the required field subject. You won't be able to upload images until all images will be in status "Ready."

How to upload photos

Choose Dataset

Select an existing dataset where images will be uploaded.

Select Filename Notation

1. Define a parsing pattern to extract metadata fields from filenames. Use predefined patterns or define your own.

2. Patterns tell the uploader how to split and interpret parts of the filename (e.g., subject ID, side, timecode).

In our example the filenames are

• t1_standard_right (2).png

• t1_standard_right (1).png

• t1_standard_left (2).png

• t1_standard_left (1).png

• etc.

So the pattern is:

<2=timeCode>_<light>_<side> (<subject>)

Select images to upload

• Drag and drop images/folders or browse files manually.

• The system validates and applies parsing rules automatically and set status for each image.

• All images that matched the pattern will be marked as "Matched" and "Ready."

  
• All images that don't matched the pattern will be marked as "Not matched" and "Not ready." In this case check the pattern and filenames.

  
• Check the detected data: subject, side and additional metadata

  

Sort timecodes as needed

Sort detected timecode in right order

Upload images

Once all images are in status "Ready", you can upload them. Click on the "Upload" button to do so.

Set fields manually

If needed, you can set field values manually. To do so:

• Select images in the table using checkboxes.

• Click "Edit" in the top-right corner of the table.

• Set field values manually and click "Save."