Neural Network Spectrogram Dimension

1. What is dimension in neural network?
2. What is dimension in CNN?
3. What's wrong with CNNS and spectrograms for audio processing?
4. What is spectrum and spectrogram?

What is dimension in neural network?

The loss surface of a neural network is a function of each one of its parameters, so the "dimension" is the number of parameters (weights, biases) in the model.

What is dimension in CNN?

The layers of a CNN have neurons arranged in 3 dimensions: width, height and depth. Where each neuron inside a convolutional layer is connected to only a small region of the layer before it, called a receptive field.

What's wrong with CNNS and spectrograms for audio processing?

Sounds are “transparent”

One challenge posed in the comparison between visual images and spectrograms is the fact that visual objects and sound events do not accumulate in the same manner. To use a visual analogy, one could say that sounds are always “transparent” [4] whereas most visual objects are opaque.

What is spectrum and spectrogram?

A spectrogram gives a running display of a sound signal as it occurs in real time; a spectrum, on the other hand, gives us a snapshot of the sound at a specific point in time. A spectrum can enable you to see, for example, the energy distribution over the different frequencies of a single vowel, like [i].