Image Labeling

Image Labeling In Action

With an Image Labeling model, you can identify the contents of an image or each frame of live video. Each prediction returns a set of labels as well as a confidence score for each label. Image Labeling can recognize people, places, and things.

If you need to know what objects are in an image, and where they are, consider using Object Detection instead.

Custom Training Models for Image Labeling

You can train a custom model that is compatible with the Image Labeling API by using Quickstart: Use Fritz AI Studio to Train a Custom Model.

Technical Specifications

Architecture Format(s) Model Size Input Output Benchmarks
MobileNet V2 variant Core ML (iOS), TensorFlow Lite (Android) ~13MB 224x224-pixel image Label + confidence score (0-100%) 38 FPS on iPhone X, 10 FPS on Pixel 2

Custom Model Compatibility Checklist

If you have a custom model that was trained outside of Fritz AI, follow this checklist to make sure it will be compatible with the Image Labeling API.

  1. Your model must be in the TensorFlow Lite (.tflite) or Core ML (.mlmodel) formats.
  2. iOS Only The name of the input layer must be named image and the output confidence.
  3. Android Only The input (image) and output layer (confidence) should be defined in the TensorFlow Lite conversion tool.
  4. The input should have the following dimensions: 1 (batch_size) x 224 (height) x 224 (width) x 3 (num_channels). Height and width are configurable.
  5. The output should have the following dimensions: 1 x number_of_labels.