Overview: Visual Recognition in Watson Studio
The IBM Watson Visual Recognition service uses deep learning algorithms to analyze images for scenes, objects, and other content. IBM Watson Studio provides a collaborative environment in the cloud where you can work with your images and your Visual Recognition custom models.
Building a visual recognition classifier
Use the graphical Visual Recognition modeler to automatically train a model to classify images for scenes, objects, or your custom content.
- Required service
- Visual Recognition service. There can only be one instance per project.
- Data format
- Image: JPEG or PNG files in a .zip file, separated by class
- Data size
- Small to medium data sets
- Building a classifier
- Each image class must contain at least 10 images, collected in a zip file
- Supported image file formats: JPEG (.jpg) and PNG (.png)
- Minimum image size: 32 x 32 pixels
- Collaborate to classify images
- Use one of five built-in models or build your own
- Test the model with sample images
- Use CoreML to develop iOS apps
- Add or remove images to retrain the model
- Use Watson Visual Recognition APIs in applications
For more information on choosing the right tool for your data and use case, see Choosing a tool.
Visual Recognition service features
Visual Recognition is a service on IBM Cloud. You can use the service in your Watson Studio projects or outside of Watson Studio.
- Built-in models
- General model - Default classification from thousands of classes
- Explicit model - Whether an image is inappropriate for general use
- Food model - Specifically for images of food items
- Text model - Text extraction from natural scene images
- Custom models - Classify images according to your specific needs
- Core ML - Include your custom models in iOS apps
- Object detection models - Train a model to detect objects within images. This feature is available as a beta.
You can find details about supported plans and features on the Visual Recognition service details page in the IBM Cloud catalog.
Watson Studio tools
Visual Recognition model builder
The Visual Recognition model builder in Watson Studio makes it quick and easy to train and test custom models. The model builder also provides quick instructions for downloading ML model (.mlmodel) files for your custom model.
Visual Recognition model builder in Watson Studio
Notebooks in Watson Studio provide a collaborative environment to use the Visual Recognition API.
Using the Visual Recognition API in a notebook in Watson Studio