Introduction to Apple’s New Initiative

Apple is set to commence utilizing images gathered for its Maps feature to train its Artificial Intelligence (AI) models. This development was revealed in a disclosure discovered by 9to5Mac, indicating that the company plans to leverage images captured for its Look Around feature, starting this month, for the additional purpose of training some of its generative AI models.

Understanding Look Around and Its Purpose

Look Around is essentially Apple’s response to Google Street View, initially introduced alongside the 2019 revamp of Apple Maps. This feature allows users to view locations from a ground-level perspective. To protect the privacy of individuals captured during these surveys, Apple ensures that faces and license plates in Look Around images are blurred.

Expansion of Apple’s Data Usage

According to the disclosure, Apple will not only use the collected images to enhance Apple Maps and improve its algorithms for blurring faces and license plates in Look Around images but also to develop and enhance other Apple products and services. The data will be pivotal in training models that power Apple products and services, including those related to image recognition, creation, and enhancement.

Apple’s Silence on the Matter

Despite the significant implications of this development, Apple did not immediately respond to requests for more information regarding this initiative.

Transparency in Data Collection

The company provides a detailed list of regions and dates for its Look Around image collection on its Apple Maps image collection policy page. Users can navigate through the list by country and then select a specific region to find out when Apple’s survey crews and vehicles are scheduled to visit their area.

Existing AI-Driven Features by Apple

Apple currently offers several features that rely on image generation models. For instance, Image Playground allows owners of Apple Intelligence-compatible devices to create new images based on text prompts. Another feature, Clean Up in Photos, enables users to remove objects from their photographs.

Precedence by Google

It’s worth noting that Google has been using Street View images to train AI models for several years. A notable example includes a 2017 project where researchers trained a machine learning model to generate professional-looking photographs from Street View imagery, showcasing the potential of leveraging such datasets for AI model training.