Dropout can certainly be applied generally - it's useful as a regularization technique (especially in wide and deep networks) to combat overfitting in other fie

Dropout can certainly be applied generally - it's useful as a regularization technique (especially in wide and deep networks) to combat overfitting in other fields than computer vision. CNNs can be used in other fields as well.