The subset of depthwise convolutional kernels from progressively deeper layers of the image pathaway Minape
Layer 1
Layer 2
Layer 3
Layer 4
Layer 5
Layer 6
Layer 7
Layer 8
Layer 9
Figure: The subset of depthwise convolutional kernels from progressively deeper layers of the image pathaway.
Figure: The total weights of the patch embedding layers in a Minape with a patch of 8 are visualized. While these layers essentially act as crude edge detectors, the industrial nature of the Mudestreda dataset prevents any discernible patterns from emerging. Interestingly, a number of filters bear a striking resemblance to noise, indicating the potential requirement for increased regularization.
The subset of depthwise convolutional kernels from progressively deeper layers of the timeseries pathaway Minape
Layer 1
Layer 2
Layer 3
Layer 4
Layer 5
Layer 6
Layer 7
Layer 8
Layer 9
Figure: The subset of depthwise convolutional kernels from progressively deeper layers of the timeseries pathaway Minape.