Image embedding guide for iOS

The MediaPipe Image Embedder task lets you convert image data into a numeric representation to accomplish ML-related image processing tasks, such as comparing the similarity of two images.

The code sample described in these instructions is available on GitHub. You can see this task in action by viewing this Web demo. For more information about the capabilities, models, and configuration options of this task, see the Overview.

Code example

The MediaPipe Tasks example code is a basic implementation of an Image Embedder app for iOS. The example uses the camera on a physical iOS device to continuously embed images, and can also run the embedder on image files from the device gallery.

You can use the app as a starting point for your own iOS app, or refer to it when modifying an existing app. The Image Embedder example code is hosted on GitHub.

Download the code

The following instructions show you how to create a local copy of the example code using the git command line tool.

To download the example code:

1. Clone the git repository using the following command:

   git clone https://github.com/google-ai-edge/mediapipe-samples
   

2. Optionally, configure your git instance to use sparse checkout, so you have only the files for the Image Embedder example app:

   cd mediapipe
   git sparse-checkout init --cone
   git sparse-checkout set examples/image_embedder/ios
   

After creating a local version of the example code, you can install the MediaPipe task library, open the project using Xcode and run the app. For instructions, see the Setup Guide for iOS.

Key components

The following files contain the crucial code for the Image Embedder example application:

• ImageEmbedderService.swift: Initializes the Image Embedder, handles the model selection, and runs inference on the input data.
• CameraViewController.swift: Implements the UI for the live camera feed input mode and visualizes the results.
• MediaLibraryViewController.swift: Implements the UI for the still image input mode and visualizes the results.

Setup

This section describes key steps for setting up your development environment and code projects to use Image Embedder. For general information on setting up your development environment for using MediaPipe tasks, including platform version requirements, see the Setup guide for iOS.

Dependencies

Image Embedder uses the MediaPipeTasksVision library, which must be installed using CocoaPods. The library is compatible with both Swift and Objective-C apps and does not require any additional language-specific setup.

For instructions to install CocoaPods on macOS, refer to the CocoaPods installation guide. For instructions on how to create a Podfile with the necessary pods for your app, refer to Using CocoaPods.

Add the MediaPipeTasksVision pod in the Podfile using the following code:

target 'MyImageEmbedderApp' do
  use_frameworks!
  pod 'MediaPipeTasksVision'
end

If your app includes unit test targets, refer to the Set Up Guide for iOS for additional information on setting up your Podfile.

Model

The MediaPipe Image Embedder task requires a trained model that is compatible with this task. For more information about the available trained models for Image Embedder, see the Models section.

Select and download a model, and add it to your project directory using Xcode. For instructions on how to add files to your Xcode project, refer to Managing files and folders in your Xcode project.

Use the BaseOptions.modelAssetPath property to specify the path to the model in your app bundle.

Create the task

You can create the Image Embedder task by calling one of its initializers. The ImageEmbedder(options:) initializer accepts values for the configuration options.

If you don't need an Image Embedder initialized with customized configuration options, you can use the ImageEmbedder(modelPath:) initializer to create an Image Embedder with the default options. For more information about configuration options, see Configuration Overview.

The Image Embedder task supports 3 input data types: still images, video files and live video streams. By default, ImageEmbedder(modelPath:) initializes a task for still images. If you want your task to be initialized to process video files or live video streams, use ImageEmbedder(options:) to specify the video or livestream running mode. The livestream mode also requires the additional imageEmbedderLiveStreamDelegate configuration option, which enables the Image Embedder to deliver image embedding results to the delegate asynchronously.

Choose the tab corresponding to your running mode to see how to create the task and run inference.

Swift

import MediaPipeTasksVision

let modelPath = Bundle.main.path(
  forResource: "model",
  ofType: "tflite")

let options = ImageEmbedderOptions()
options.baseOptions.modelAssetPath = modelPath
options.quantize = true
options.l2Normalize = true

let imageEmbedder = try ImageEmbedder(options: options)
    

import MediaPipeTasksVision

let modelPath = Bundle.main.path(
  forResource: "model",
  ofType: "tflite")

let options = ImageEmbedderOptions()
options.baseOptions.modelAssetPath = modelPath
options.runningMode = .video
options.quantize = true
options.l2Normalize = true

let imageEmbedder = try ImageEmbedder(options: options)
    

import MediaPipeTasksVision

// Class that conforms to the `ImageEmbedderLiveStreamDelegate` protocol and
// implements the method that the image embedder calls once it finishes
// embedding each input frame.
class ImageEmbedderResultProcessor: NSObject, ImageEmbedderLiveStreamDelegate {

  func imageEmbedder(
    _ imageEmbedder: ImageEmbedder,
    didFinishEmbedding result: ImageEmbedderResult?,
    timestampInMilliseconds: Int,
    error: Error?) {

    // Process the image embedder result or errors here.

  }
}

let modelPath = Bundle.main.path(
  forResource: "model",
  ofType: "tflite")

let options = ImageEmbedderOptions()
options.baseOptions.modelAssetPath = modelPath
options.runningMode = .liveStream
options.quantize = true
options.l2Normalize = true

// Assign an object of the class to the `imageEmbedderLiveStreamDelegate`
// property.
let processor = ImageEmbedderResultProcessor()
options.imageEmbedderLiveStreamDelegate = processor

let imageEmbedder = try ImageEmbedder(options: options)
    

Objective-C

@import MediaPipeTasksVision;

NSString *modelPath = [[NSBundle mainBundle] pathForResource:@"model"
                                                      ofType:@"tflite"];

MPPImageEmbedderOptions *options = [[MPPImageEmbedderOptions alloc] init];
options.baseOptions.modelAssetPath = modelPath;
options.runningMode = MPPRunningModeImage;
options.quantize = YES;
options.l2Normalize = YES;

MPPImageEmbedder *imageEmbedder =
  [[MPPImageEmbedder alloc] initWithOptions:options error:nil];
    

@import MediaPipeTasksVision;

NSString *modelPath = [[NSBundle mainBundle] pathForResource:@"model"
                                                      ofType:@"tflite"];

MPPImageEmbedderOptions *options = [[MPPImageEmbedderOptions alloc] init];
options.baseOptions.modelAssetPath = modelPath;
options.runningMode = MPPRunningModeVideo;
options.quantize = YES;
options.l2Normalize = YES;

MPPImageEmbedder *imageEmbedder =
  [[MPPImageEmbedder alloc] initWithOptions:options error:nil];
    

@import MediaPipeTasksVision;

// Class that conforms to the `MPPImageEmbedderLiveStreamDelegate` protocol
// and implements the method that the image embedder calls once it finishes
// embedding each input frame.
@interface APPImageEmbedderResultProcessor : NSObject 

@end

@implementation APPImageEmbedderResultProcessor

-   (void)imageEmbedder:(MPPImageEmbedder *)imageEmbedder
    didFinishEmbeddingWithResult:(MPPImageEmbedderResult *)imageEmbedderResult
         timestampInMilliseconds:(NSInteger)timestampInMilliseconds
                           error:(NSError *)error {

    // Process the image embedder result or errors here.

}

@end

NSString *modelPath = [[NSBundle mainBundle] pathForResource:@"model"
                                                      ofType:@"tflite"];

MPPImageEmbedderOptions *options = [[MPPImageEmbedderOptions alloc] init];
options.baseOptions.modelAssetPath = modelPath;
options.runningMode = MPPRunningModeLiveStream;
options.quantize = YES;
options.l2Normalize = YES;

// Assign an object of the class to the `imageEmbedderLiveStreamDelegate`
// property.
APPImageEmbedderResultProcessor *processor =
  [APPImageEmbedderResultProcessor new];
options.imageEmbedderLiveStreamDelegate = processor;

MPPImageEmbedder *imageEmbedder =
  [[MPPImageEmbedder alloc] initWithOptions:options error:nil];
    

Configuration options

This task has the following configuration options for iOS apps:

When the running mode is set to livestream, the Image Embedder requires the additional imageEmbedderLiveStreamDelegate configuration option, which enables the Image Embedder to deliver image embedding results asynchronously. The delegate must implement the imageEmbedder(_:didFinishEmbedding:timestampInMilliseconds:error:) method, which the Image Embedder calls after processing the results of embedding each input image frame.

Prepare data

You need to convert the input image or frame to an MPImage object before passing it to the Image Embedder. MPImage supports different types of iOS image formats, and can use them in any running mode for inference. For more information about MPImage, refer to the MPImage API.

Choose an iOS image format based on your use case and the running mode your application requires.MPImage accepts the UIImage, CVPixelBuffer, and CMSampleBuffer iOS image formats.

UIImage

The UIImage format is well-suited for the following running modes:

• Images: images from an app bundle, user gallery, or file system formatted as UIImage images can be converted to an MPImage object.

• Videos: use AVAssetImageGenerator to extract video frames to the CGImage format, then convert them to UIImage images.

// Load an image on the user's device as an iOS `UIImage` object.

// Convert the `UIImage` object to a MediaPipe's Image object having the default
// orientation `UIImage.Orientation.up`.
let image = try MPImage(uiImage: image)
    

// Load an image on the user's device as an iOS `UIImage` object.

// Convert the `UIImage` object to a MediaPipe's Image object having the default
// orientation `UIImageOrientationUp`.
MPImage *image = [[MPPImage alloc] initWithUIImage:image error:nil];
    

The example initializes an MPImage with the default UIImage.Orientation.Up orientation. You can initialize an MPImage with any of the supported UIImage.Orientation values. Image Embedder does not support mirrored orientations like .upMirrored, .downMirrored, .leftMirrored, .rightMirrored.

For more information about UIImage, refer to the UIImage Apple Developer Documentation.

CVPixelBuffer

The CVPixelBuffer format is well-suited for applications that generate frames and use the iOS CoreImage framework for processing.

The CVPixelBuffer format is well-suited for the following running modes:

• Images: apps that generate CVPixelBuffer images after some processing using iOS's CoreImage framework can be sent to the Image Embedder in the image running mode.

• Videos: video frames can be converted to the CVPixelBuffer format for processing, and then sent to the Image Embedder in video mode.

• livestream: apps using an iOS camera to generate frames may be converted into the CVPixelBuffer format for processing before being sent to the Image Embedder in livestream mode.

// Obtain a CVPixelBuffer.

// Convert the `CVPixelBuffer` object to a MediaPipe's Image object having the default
// orientation `UIImage.Orientation.up`.
let image = try MPImage(pixelBuffer: pixelBuffer)
    

// Obtain a CVPixelBuffer.

// Convert the `CVPixelBuffer` object to a MediaPipe's Image object having the
// default orientation `UIImageOrientationUp`.
MPImage *image = [[MPPImage alloc] initWithUIImage:image error:nil];
    

For more information about CVPixelBuffer, refer to the CVPixelBuffer Apple Developer Documentation.

CMSampleBuffer

The CMSampleBuffer format stores media samples of a uniform media type, and is well-suited for the livestream running mode. Live frames from iOS cameras are asynchronously delivered in the CMSampleBuffer format by iOS AVCaptureVideoDataOutput.

// Obtain a CMSampleBuffer.

// Convert the `CMSampleBuffer` object to a MediaPipe's Image object having the default
// orientation `UIImage.Orientation.up`.
let image = try MPImage(sampleBuffer: sampleBuffer)
    

// Obtain a `CMSampleBuffer`.

// Convert the `CMSampleBuffer` object to a MediaPipe's Image object having the
// default orientation `UIImageOrientationUp`.
MPImage *image = [[MPPImage alloc] initWithSampleBuffer:sampleBuffer error:nil];
    

For more information about CMSampleBuffer, refer to the CMSampleBuffer Apple Developer Documentation.

Run the task

To run the Image Embedder, use the embed() method specific to the assigned running mode:

• Still image: embed(image:)
• Video: embed(videoFrame:timestampInMilliseconds:)
• Livestream: embedAsync(image:timestampInMilliseconds:)

The following code samples show basic examples of how to run Image Embedder in these different running modes:

Swift

let result = try imageEmbedder.embed(image: image)
    

let result = try imageEmbedder.embed(
  videoFrame: image,
  timestampInMilliseconds: timestamp)
    

try imageEmbedder.embedAsync(
  image: image,
  timestampInMilliseconds: timestamp)
    

Objective-C

MPPImageEmbedderResult *result =
  [imageEmbedder embedImage:image error:nil];
    

MPPImageEmbedderResult *result =
  [imageEmbedder embedVideoFrame:image
           timestampInMilliseconds:timestamp
                             error:nil];
    

BOOL success =
  [imageEmbedder embedAsyncImage:image
           timestampInMilliseconds:timestamp
                             error:nil];
    

The Image Embedder code example shows the implementations of each of these modes in more detail embed(image:), embed(videoFrame:timestampInMilliseconds:), and embedAsync(image:timestampInMilliseconds:). The example code allows the user to switch between processing modes which may not be required for your use case.

Note the following:

• When running in video mode or livestream mode, you must also provide the timestamp of the input frame to the Image Embedder task.

• When running in image or video mode, the Image Embedder task blocks the current thread until it finishes processing the input image or frame. To avoid blocking the current thread, execute the processing in a background thread using iOS Dispatch or NSOperation frameworks. If your app is created using Swift, you can also use Swift Concurrency for background thread execution.

• When running in livestream mode, the Image Embedder task returns immediately and doesn't block the current thread. It invokes the imageEmbedder(_:didFinishEmbedding:timestampInMilliseconds:error:) method with the results, after embedding each input frame. The Image Embedder invokes this method asynchronously on a dedicated serial dispatch queue. For displaying results on the user interface, dispatch the results to the main queue after processing the results. If the embedAsync function is called when the Image Embedder task is busy processing another frame, the Image Embedder ignores the new input frame.

Handle and display results

Upon running inference, the Image Embedder returns an ImageEmbedderResult object that contains a list of embeddings (either floating point or scalar-quantized) for the input image.

The following shows an example of the output data from this task:

ImageEmbedderResult:
  Embedding #0 (sole embedding head):
    float_embedding: {0.0, 0.0, ..., 0.0, 1.0, 0.0, 0.0, 2.0}
    head_index: 0

This result was obtained by embedding the following image:

You can compare the similarity of two embeddings using the ImageEmbedder.cosineSimilarity function.

let similarity = try ImageEmbedder.cosineSimilarity(
  embedding1: result.embeddingResult.embeddings[0],
  embedding2: otherResult.embeddingResult.embeddings[0])
    

NSNumber *similarity = [MPPImageEmbedder
      cosineSimilarityBetweenEmbedding1:result.embeddingResult.embeddings[0]
                          andEmbedding2:otherResult.embeddingResult.embeddings[0]
                                  error:nil];