Kaggle Tensorflow Speech Recognition Challenge – Towards Data Science

The training data supplied by Google Brain consists of ca. Only 10 of these are classes you need to identify, the others should go in the'unknown' or'silence' classes. There are a couple of things you can do to get a grip on the data you're working with. This data set is not completely cleaned up for you. For example, some files are not exactly 1 second long. And there are no'silence' files as such.