Need to understand, moderate, or automate responses to lots of text? How about recognizing text in images and videos? Welcome to the one platform where you can power smart applications using videos, images, text and more.
One platform for all of your AI needs
Clarifai has been a leader in computer vision AI since the beginning. Now we're introducing powerful text classification and visual text recognition models that seamlessly integrate language understanding with computer vision on our platform. With our new Natural Language Processing (NLP) and Visual Text Recognition technology, we've augmented our AI platform so that you can detect, understand and classify blocks of text to extract meaning.
NLP complements Computer Vision in a wide range of scenarios. Social media, product reviews and user-generated content of all kinds require a complementary understanding of text and visual data, as users upload their images and write their comments.
Clarifai now makes it easier to have all your AI in one place. Plus, text and visual recognition technologies can be mixed and matched within multiple workflows, so that you can build AI solutions that have never been possible before.
Meet our new models
Text Classification
You can think of our Text Classification model as a general-purpose model for language understanding. Any topic, category, classification or distinction that you can think of can be a custom concept for your model.
Want to track and understand consumer sentiment? Create "Positive", "Neutral" and "Negative" concepts, and train your model on real customer feedback. How about an auto-reply chatbot for your website? Train a model to understand common support issues and provide customers with useful guidance based on their questions.
You can even understand and organize documents and public records. Or auto-classify news articles. Or even track industry trends.
Whatever your use case, if it involves written language, Text Classification can help you solve business problems in new ways. Operate at a scale in a way that was never possible with a human-only workforce.
Text Moderation
Text Moderation helps you flag obscene and indecent text posted by users. Moderation is such a common use case in text analysis, that our researchers have developed a fast and lightweight model with custom architecture that is specifically designed with content moderation in mind.
Text Moderation is ready to use right out of the box with no custom training required. It'll automatically predict content that is:
- Toxic
- Severely toxic
- Obscene
- Threatening
- Insulting
- Hate speech
Text Moderation can also be trained to understand custom concepts, so that you can address specific moderation concerns relevant to your user base.
Visual Text Recognition
Visual Text Recognition identifies text that appears in your image or video data so that you can "understand" this text with AI. Our new model can be used to extract text from a scene such a billboard image, embedded text within a meme, or a document that you've scanned. Visual Text Recognition lets you "digitize" text so that it can be edited, searched, stored, displayed and analyzed.
Multimodal Workflows
Clarifai is your single end-to-end AI platform for consolidating all your AI modeling projects. This means you can network models to have them work together to solve your business problems. For example user generated content includes image and text data and being able to moderate it in unison provides more accurate coverage. Now you can understand the meaning of text within the context of an image and vice-versa.
Start building
To learn more about how to use text models for your business application, take a look at these walkthroughs:
