Meta’s new AI model can detect objects within photos and videos

Updated : Apr 07, 2023 13:07
|
Editorji News Desk

Meta has created an AI model called "Segment Anything," which can identify objects in images and videos without prior training, even detecting items not listed in the original data set.

This eliminates the need for pre-existing data sets and enables more precise object recognition, reducing the requirement for further training, which can increase efficiency.

However, the model may experience slow performance with complex image processing, and specialised AI tools are still needed for specific domains.

Meta will make both the AI model and dataset available for download under a non-commercial license, promoting research and enhancing accessibility to technology.

Meta

Recommended For You

editorji | Tech

OnePlus 15R review: A clear shift in what the R-series stands for

editorji | Tech

POCO C85: A strong package with some rough edges

editorji | Tech

Lenovo Legion 27Q10 Review: A serious gaming monitor under ₹20,000

editorji | Tech

Redmi 15C 5G Review: Xiaomi’s budget phone that just refuses to die in a day

editorji | Tech

Samsung Galaxy M17 5G Review: Best Budget Samsung Phone Under ₹15,000?