With its latest update for Google Photos, the search engine giant lets its users label the pictures on the app. This way, the company will be able to train and improve its object recognition algorithms.
This is being advertised as a virtuous cycle of AI development, which will help make Google photos more useful. It’s the same reason that CAPTCHAs ask you to identify cars and motorbikes in images. To make AI algorithms intelligent, they have to be trained by humans. Hence, this practice is not unusual. A vast majority of such applications are based on data labeled by users.
This feature is available in the most recent version of Google photos. To activate it, follow these steps:
- Tap on the search button in the app’s menu.
- Select “Help improve Google Photos” from the scroll menu.
- You will be presented with four options:
- Describe your printing preferences for photos
- Describe your preferred collages or animations
- Identify which photos belong to which holiday events (e.g., Christmas or Halloween
- Identify the contents of photos (“Name the most important things in this photo”)
- Select either of the options and label your image.
According to Google,
It may take time to see the impact your contributions have on your account, but your input will help improve existing features and build new ones; for example, improved suggestions on which photos to print or higher quality creations that you would like. You can delete your answers at any time.
This feature is not exactly new though, its underlying software is much older and is powered by “Crowdsource by Google,” which was launched by the company in 2016.