Friday, 24 March, 2023

MIT-IBM developed a faster way to train video recognition AI

<p>A team from the MIT &ndash; IBM Watson AI Lab has come up with a solution to make the video &ndash; recognition machine learning models more efficient by reducing the size of the model&rsquo;s size. This model also aims to speed up training and improve performance on mobile devices. Researches have done this by shifting video recognition models view time and designing a temporal shift module.</p>