See What Objects Facebook Has Identified In Your Photos
Facebook has built-in face recognition that works on the photos you upload and share. When you upload a photo, the face recognition suggests which friends you should tag in it. Facebook doesn’t just recognize the people in photos; it also recognizes objects in your photos. The photos we upload to Facebook aren’t all selfies; food and cats are also popular subjects in our photos. Facebook can and does recognize these things and duly adds tags to it that we can’t see. If you want to check out what objects Facebook has identified in photos, you need the Show Facebook Computer Vision Tags extension for Chrome.
Show Facebook Computer Vision Tags isn’t available in the Chrome Web store so you can’t install it the traditional way. Download it as a zipped file first and extract it.
Go to chrome://extensions/ and enable ‘Developer Mode’. Next, click on ‘Load unpacked extension…’ and select the folder you just extracted.
The extension will be installed instantly. Show Facebook Computer Vision Tags doesn’t need to connect to your Facebook account. It’s simply reading meta data that is attached to the photos by Facebook when it loads in Chrome. Start scrolling through your news feed.
You will notice little tags on photos that indicate how many people or what object is in them. Facebook will almost always, without fail, recognize a picture of food. For most other things, it’s not as good. The tags can include what time of day it is, whether or not there is text in the photo, what activity is happening, and if it’s an indoor photo or an outdoor photo, among other things.
Not all photos have these tags. When Facebook is unable to identify items in it, it doesn’t add a tag to it. It seems it only analyzes the photos when they are uploaded and doesn’t go back to run the recognition algorithm again.
Show Facebook Computer Vision Tags is a Chrome extension but as per the developer, you should be able to run it in the Dev version of Firefox which has support for Chrome extensions.