From the course: Digital Marketing Trends

Unlock the full course today

Join today to access over 23,200 courses taught by industry experts.

Getting a handle on multimodal AI

Getting a handle on multimodal AI

From the course: Digital Marketing Trends

Start my 1-month free trial Buy for my team

Getting a handle on multimodal AI

“

- [Instructor] Among the many areas where human intelligence is different from machines is our ability to describe the images we see or paint a picture based on a story we read or heard. Artificial intelligence hasn't been able to perform that task at all until very recently. That's where multimodal learning comes in. So what is multimodal learning and how does it work? Let's step back to see the bigger picture by considering the two types of unstructured data used by AI. On one hand, there's natural language processing where an AI finds patterns in the words we hear, read, and say, and then there's computer vision which is how a machine sees the world. Multimodal learning integrates the two data types in a way that grasps the relationship between words and the images they represent. From the outside it operates in a simple way, give a multimodal AI a text prompt and it produces a variety of images based on what you just…

Contents