Facebook AI Lumos Can Find Your Photos Even Those You Are Not Tagged In [VIDEO]

• comments
• 
• 

Facebook is relying on Artificial Intelligence of Lumos to scan photos uploaded on the website. Whenever people search for photos, Lumos will perform image recognition techniques and queues up the best search matching with the keywords input given by the user. By doing this, Facebook leapfrogs from the traditional text-based search technology to image-driven technology.

According to the Facebook blog posted by Facebook's Director of Applied Machine Learning, Joaquin Quinonero Candela said, "We've pushed computer vision to the next stage with the goal of understanding images at the pixel level. This helps our systems do things like recognize what's in an image, what type of scene it is, if it's a well-known landmark, and so on". He asserted that days are gone by when images are ranked based on tags and caption.

Facebook's Artificial Intelligence platform kicked off as a small project and now developed as a full-scale production tool and a computer vision. When images were passed through Lumos, it identifies images and scenes and that in turn adds value to the search.

Facebook has 12 preset actions that any photo typically contains like "people walking", "people riding horses", "people dancing", people playing instruments" and etc. The cutting-edge deep learning techniques process billions of photos to understand the photo content through image recognition and image embedding, before prioritizing and ranking them based on their semantic meaning.

Facebook has empowered its technical teams to run through a large chunk of data amassed with it. Dozens of teams are working on training and deploying Lumos with 200 visual models for the purpose of objectionable content detection, spam detection and auto-image captioning. As per Tech Crunch, Facebook has launched Automatic Alternative Tool for visually impaired persons to recognize the contents of the image through the innovative text-to-speech tool.

Facebook teams have worked and gained a lot of experience on Lumos, but the road ahead is very long and exciting. In future, Lumos will plunge into video and immersive formats to help in Facebook in achieving new real-time possibilities and product experiences.