Salesforce Research uses computer vision combined with language and audio to solve real problems: detecting logos on social media, counting products on retail shelves, and classifying cancer in medical images. The team also uses drones to spot sharks off California coasts.
December 5, 2019.
100,000 plus organizations.
Computer vision is a field of understanding natural images as well as video. A couple of applications are, for instance, trying to find your company logo on social media or counting specific types of products in supermarket shelves. And medical computer vision, we classify different types of cancer, for instance, in a variety of different modalities.
Human knowledge it starts from vision. So you see something, you use some words to express what you see and then you start to imagine some pictures. To build computer vision system like a human, it will change everything.
Combining language and vision and audio and integrating those different types of information can help us to optimize all these different types of data.
We are trying to do some volunteer work by integrating our AI technologies. For instance, they are working on a project. We are trying to protect people from shark attacks along the shoreline of California. We are using drones and computer vision techniques to detect sharks, to prevent tourists from being attacked by them, these kind of wild animals. This kind of work is non profit. We care about our neighbors, our communities here.
The most exciting things about Salesforce Research is that Salesforce is a platform company for us. It's never enough to just solve one problem and then release one app when we solve an AI problem. Now 100,000 plus organizations can use these tools to solve their own problems. And that to me is a much more exciting goal than just trying to solve and build a single application. Sam.
