After transforming timeseries into an image format, I get a width-height ratio of ~135. Typical image CNN applications involve either square or reasonably-rectangular proportions - whereas mine look nearly like lines:
(16000, 120, 16) = (width, height, channels).
Are 2D CNNs expected to work well with such aspect ratios? What hyperparameters are appropriate - namely, in Keras/TF terms,
kernel_size (is 'unequal' preferred, e.g.
strides=(16, 1))? Relevant publications would help.
width == timesteps. The images are obtained via a transform of the timeseries, e.g. Short-time Fourier Transform.
channels are the original channels.
height is the result of the transform, e.g. frequency information. The task is binary classification of EEG data (w/ sigmoid output).