Visual Recognition understands the contents of images - visual concepts tag the image, find human faces, approximate age and gender, and find similar images in a collection. You can also train the service by creating your own custom concepts. Use Visual Recognition to detect a dress type in retail, identify spoiled fruit in inventory, and more.
Analyze images for scenes, objects, faces, text, and other subjects that can give you insights into your visual content.
Developers can use the Visual Recognition service for any task that requires understanding images automatically or at scale. By categorizing and finding similar images for you, this service eliminates the need to manually classify images one-by-one. Use Visual Recognition to simplify and expedite tasks such as:
Organize image libraries into categorized folders
Segment user interests from social media pictures
Find high-quality images with specific content faster
Recognize custom content from images
Identify matching images
Image or website URLs
Custom: Classifier name, JPEG images (positive examples for each class, and negative examples)
Custom: Collections - your own images that you want to search for similar images
Face detection (Gender, age range, celebrity (very limited see docs)
Question: What is the difference between a "Class" and "Classifier" for pricing purposes in Visual Recognition? Answer: Using the dog breed scenario in the demo as an example, each dog breed, such as "Husky", is a class. The classifier is "Dog Breed" which is the collection of classes. Learn more about using your own data.
Question: How many classes can I create for my one free classifier? Answer: However many classes you want. You are limited to 5000 free training images and one classifier. For best results, use a minimum of 50 images per class. Learn more about using your own data.