Computer Vision News - October 2020

Snorkel.AI 19 powerful feedback loop. This means you can see results and improvements much faster and more efficiently, and significantly lower the bar in terms of the level of expertise people need to be able to contribute to machine learning. “We see this as being a dramatic change to the way that business is done in the coming decade ,” he asserts. “In the same way that cloud computing has changed the software development landscape in the last 10 years, machine learning is poised to do the same thing, but even more so I’d say. Now anyone with subject matter expertise should be able to develop state-of-the-art machine learning techniques.With Snorkel, all you need to bring to the table is that domain expertise and we can give you access to state-of-the-art models so that you can make an impact in your field.” Thinking about medical imaging in particular, Paroma reveals the team have worked with the Stanford radiology lab and doctors at Stanford Hospital . “One of our most recent collaborations on the research side was with a cardiac pediatrician where we used Snorkel to label thousands of heart MRIs for children,” she tells us. “We had a publication in Nature Communications about that. We also had a collaborationusing Snorkel for medical device surveillance. We are