Can Convolutional Neural Networks be applied in domains other than image recognition?

Question

I'm new in this argument, my question is:

Can convolution be applied in other contexts different from image recognition? Is there a good source to learn from?

Answer 1

Yes, Convolutional Neural Networks(CNNs) can and have been applied to non-image problems. Arguably, any problem in which the location of a feature(s) is relevant can be attempted via CNNs. CNNs works under the assumption that points close to each other in the data share some correlations/relationship whist points further apart don't share as much information. So, theoretically, if you can phrase your problem so that it meets this requirement(s), it can be attempted by a convolutional neural network. Here are few applications of CNNss that are dont involve images:

• Speech recognition
• Character-level Convolutional Networks for Text Classification
• Text classification using CNNs
• Natural Language Processing
• Time Series Forecasting
• Sentence modeling

• I've used 1-D CNNs on spectral data. Here are some examples of CNNs applied to spectral data :

  • Spectral Convolutional Neural Network for music classification
  • Convolutional neural network based classification for hyperspectral data

Answer 2

CNNs are generally designed for 2D (generally image) data. Hence, other usages are most likely a "hack" to CNN logic.

You can check how to extract features and classify texts with CNN through here.