Bing Chat Will Present The GPT-4-Based Image Recognition Feature

JAKARTA - Soon Microsoft is claimed to be launching an image recognition feature on Bing Chat. Users can identify images with chatbots.

This feature will work when the user uploads an image that can then ask a question to the Bing Chat regarding the image.

Armed with OpenAI's GPT-4 vision model, the image recognition feature is able to detect and understand objects in the image.

A bit over 10%. Not on Mobile yet, no.

Not only that, image recognition can also provide detailed explanations about images using real-life examples.

Similar to Google Lens, the Microsoft image recognition feature is still in the testing phase to some desktop users, but plans to be extended to Bing Chat mobile users in a few weeks.

Microsoft's CEO of Advertising and Web Service, Mikhail Parakhin recently stated on his Twitter that the testing phase is currently limited to 10 percent, up from 5 percent last week.

After its launch, Bing Chat users will see a new option next to the voice icon. There, they can upload the image directly to the chat or paste the image from the internet and ask Bing to explain it.

Image recognition features are also included in Windows Copilot. Users can drag images from the desktop or File Explorer, drop them to Copilot, and ask AI to explain the image or create something similar.

Users can also copy images directly into PowerPoint, Word, or billboards. Before being launched to the public, Parakhin said his company wanted to ensure safety and maintain the quality of regular chat interactions first.

So, as confidence in quality and safety increases, this feature will be available for a wider user base soon. This was quoted from Windwos Latest and ONMSFT, Monday, June 26.