ylliX - Online Advertising Network
Gaze Link Wins Best Android App in Gemini API Developer Competition

Gaze Link Wins Best Android App in Gemini API Developer Competition



Posted by Thomas Ezan – Sr Developer Relation Engineer (@lethargicpanda)

We’re excited to announce Gaze Link as the winner of the Best Android App for our Gemini API Developer Competition!

This innovative app demonstrates the potential of the Gemini API in providing a communication system for individuals with Amyotrophic lateral sclerosis (ALS) who develop severe motor and verbal disabilities, enabling them to type sentences with only their eyes.

About Gaze Link

Gaze Link uses Google’s Gemini 1.5 Flash model to predict the user’s intended sentence based on a few key words and the context of the conversation.

For example if the context is “Is the room temperature ok?” and the user replies “hot AC two” the app will leverage Gemini to generate the full sentence “I am hot, can you turn the AC down by two degrees?”.

The Gaze Link team took advantage of Gemini 1.5 Flash multilingual capabilities to let the app generate sentences in English, Spanish and Chinese, the three languages currently supported by the app.

We were truly impressed by the Gaze Link app. The team used the Gemini API combined with ML Kit Face Detection to empower individuals with ALS providing them with a powerful communication system that is both accessible and affordable.

With Gemini 1.5 Flash currently supporting 38 languages, it is possible for Gaze Link to add support for more languages in the future. In addition, the model’s multimodal abilities could enable the team to enhance the user experience by integrating image, audio and video to augment the context of the conversation.

Build with the Gemini API

The result of the integration of the Gemini API in Gaze Link is inspiring. If you are working on an Android app today, we encourage you to learn about the Gemini API capabilities to see how you can successfully add generative AI to your app and delight your users.

To get started, go to the Android AI documentation!




Source link

Leave a Reply

Your email address will not be published. Required fields are marked *