Deep Learning and Speech Recognition - What's the Future of Machine Learning?

The level of perceptibility of the adversarial perturbation is still kind of high. It's become possible to make sounds that sound like normal speech, but are actually interpreted as a different sentence than the human hears. There isn't a lot of point in trying to summarize exactly which architecture and which learning approach got to which level of performance. Some core ideas from the 1980s are still used today.