Object Detection Call

Object Detection

Object Detection calls are calls that can take an image and identify objects within it, returning their labels, confidence scores, and bounding box coordinates.

Example normalized Object Detection call

The following is an example of sending the file street_photo.jpg into Huggingface with the facebook/detr-resnet-50 model and getting the detected objects back in form of an array of ObjectDetectionItem.

use Drupal\ai\OperationType\GenericType\ImageFile;
use Drupal\ai\OperationType\ObjectDetection\ObjectDetectionInput;

$image_binary = file_get_contents('street_photo.jpg');
$normalized_file = new ImageFile($image_binary, 'image/jpeg', 'street_photo.jpg');
$input = new ObjectDetectionInput($normalized_file);
/** @var array $detection_items */
$detection_items =  \Drupal::service('ai.provider')->createInstance('huggingface')->objectDetection($input, 'facebook/detr-resnet-50', ['my-custom-call'])->getNormalized();

foreach ($detection_items as $item) {
  // Output each detected object with its bounding box.
  print 'Found ' . $item->getLabel() . ' with confidence ' . $item->getConfidenceScore() . "\n";
  print 'Bounding box: ' . json_encode($item->getBoundingBox()) . "\n";
}

Object Detection Interfaces & Models

The following files defines the methods available when doing a Object Detection call as well as the input and output.

• ObjectDetectionInterface.php
• ObjectDetectionInput.php
• ObjectDetectionOutput.php

Object Detection Explorer

If you install the AI API Explorer, you can go configuration > AI > AI API Explorer > Object Detection Explorer under /admin/config/ai/explorers/object_detection_generator to test out different calls and see the code that you need for it.