July 20, 2020

Clarifai Release 6.4

Table of Contents:

We are proud to introduce Clarifai version 6.4, which seamlessly integrates AI for text with AI for images within our platform. 

Banner of  of Clarifai release 6.4

Multi-Modal Workflows

We are building a complete platform for the end-to-end AI lifecycle. With multi-modal work flows you can combine multiple "modes" of AI into one application. Computer vision and natural language processing can be combined in one solution - an important capability in a world where almost all all communication is a combination of images, video and text. The fact that it all happens on one platform means that these different "modes" work together seamlessly. 

Text Embedding Model

Text models can be trained to understand the meaning of text passages. Our text embedding model can be trained on any set of concepts that you can imagine. Since language itself is so versatile, text classifications can be pretty versatile too. You can train a model to give you valuable social media metrics, categorize academic research papers, or even power a custom chatbot.

Text Moderation Model

Toxic, insulting, obscene and threatening content. The world is full of it, unfortunately. Use this model to moderate user generated content so that you can protect your brand, and deliver a positive customer experience online. Our text moderation is a lightweight and fast model that is ready to use out of the box. 

Visual Text Recognition

Visual text recognition helps you convert printed text in images and videos into machine-encoded text. You can input a scanned document, a photo of a document, a scene-photo (such as the text on signs and billboards), or text superimposed on an image (such as in a meme) and output the words and individual characters present in the images. VTR lets you "digitize" text so that it can be edited, searched, stored, displayed and analyzed.

CSV Uploads

You can now uploads your inputs, labels and metadata as CSV files. This means that you can use your favorite text editor, or spreadsheet software to configure and label your training data. This is very helpful when working with text inputs, or images/videos that you are using for classification models. 

Read the 6.4 Changelog.