Okay, so they've taken an old method and applied it to deep learning and they say it's better.

But I'm not seeing anything beyond that in this article.

I haven't read the linked paper yet, but I'm not seeing much meat in this article. Just giving us the abstract from the paper might have been better.