Google has recently open-sourced their model for identifying the images in a photo. Boasting a 93.9% accuracy rate, the latest version is their best so far.
The tool takes previous experience with similar photos to identify what is happening and who or what is in the picture.
It seems similar to Facebook’s artificial intelligence photo descriptions and promises to get more accurate over time.
Read more about this new project in this article and how it might be beneficial for blind and visually impaired users.
Have you found Facebook’s photo descriptions helpful? Share your thoughts in the comment section.
What’s in that photo? Google open-sources caption tool in TensorFlow can tell you