RectLabel

How to use? Read our Help page.

Post the problem to our Github issues.

Have questions? Send an email to support@rectlabel.com.

Core ML menus

• Load Core ML model
• Auto labeling
• Auto text recognition for lines and words
• Toggle auto process
• Search incorrectly predicted images

Load Core ML model

• Apple provides Core ML Models.
• To convert your ML model to the Core ML format, use coremltools.
• You can convert YOLOv5, YOLOv8/YOLO11, and MobileNetV2 + SSDLite models to Core ML models.

Auto labeling

Auto label images using the loaded Core ML model.

• You can change the confidence threshold.
• if the model does not include the non-maximum suppression, you can change the overlap threshold.

View menu.

• “Show labels on boxes” to show confidences on boxes.

Settings menus.

• “Skip backup dialog when overwrite” is to skip the backup dialog when overwrite.
• “Clear existing labels when auto label all images” is to clear existing labels when auto label all images.
• “Save confidence values when YOLO format” is to save confidence values when YOLO format.

Auto text recognition for lines and words

Using Vision Framework, automatic text recognition is performed.

• You can change the language.
• You can change the type to lines or words.

Toggle auto process

You can auto label when change the image if you check on the “Auto process” checkbox on the Core ML dialog.

Search incorrectly predicted images

Show training images which predicted labels and boxes are incorrect compared with the current annotations. Such as “Not detected”, “Not classified”, and “Falsely detected”.