This documentation is for AutoML Vision, which is different from Vertex AI. If you are using Vertex AI, see the Vertex AI documentation.

AutoML Vision Edge terminology

Term Description
Android Studio The official integrated development environment (IDE) for Android app development.
CocoaPods A dependency manager used here for iOS app development; for Swift and Objective-C Cocoa projects to provide a standard format for managing external libraries.
Core ML A machine learning framework used in Apple products. TensorFlow lite models can be converted to CoreML format for use on Apple devices.
Container ("export to Docker/container") The runtime instance of an image; one of the export options for your model using AutoML Vision Edge.
Edge devices A device that provides compute capability outside the cloud. The demand for privacy/confidentiality, low latency and bandwidth constraints drive demand for predictions with our models on these devices. Compute and power constraints lead to models specialized for them.
Edge TPU A type of Edge device; Google's purpose-built application-specific integrated circuit (ASIC) designed to run inference at the edge. Supports .tflite models only.
Firebase A mobile and web application development platform.
FlatBuffers Similar to Protocol buffers, with the primary difference being that FlatBuffers do not need a parsing/ unpacking step to a secondary representation before you can access data, often coupled with per-object memory allocation.
IoT Internet of Things (IoT); the use of network-connected devices, embedded in the physical environment, to improve some existing process or to enable a new scenario not previously possible.
ML Kit ML Kit acts as an API layer to your custom model; a mobile software development kit (SDK) that allows you to use a custom model on-device.
Pillow The Python Imaging Library (PIL) adds image processing capabilities to your Python interpreter; Pillow is a modified version of the base PIL.
Protocol buffers ("protobuf") Google's language-neutral, platform-neutral, extensible mechanism for serializing structured data. Similar to FlatBuffers.
TensorFlow TensorFlow is an end-to-end open source platform for machine learning; software used to create a machine learning model.
TensorFlow lite model (TF Lite/model.tflite) A TensorFlow ML model that has been compressed for use on mobile and embedded devices.

  • TF Lite converter - TensorFlow Lite uses the optimized FlatBuffer format to represent graphs. Therefore, a TensorFlow model (protocol buffer) needs to be converted into a FlatBuffer file before deploying to clients.
  • TF Lite interpreter - A class that does the job of a tf.Session(), only for TF Lite models as opposed to regular TensorFlow models.
tf.session() A class for running TensorFlow operations using a TensorFlow model.
Xcode Xcode is an integrated development environment (IDE) for macOS containing a suite of software development tools developed by Apple for developing software for macOS, iOS, watchOS, and tvOS.